Professional Documents
Culture Documents
طرق التنبؤ الإحصائي
طرق التنبؤ الإحصائي
Uploaded by
الاقتصاد امين الدرباقCopyright
Available Formats
Share this document
Did you find this document useful?
Is this content inappropriate?
Report this DocumentCopyright:
Available Formats
طرق التنبؤ الإحصائي
طرق التنبؤ الإحصائي
Uploaded by
الاقتصاد امين الدرباقCopyright:
Available Formats
ﺟﺎﻣﻌﺔ ﺍﻟﻤﻠﻚ ﺳﻌﻮﺩ
ﻗﺴﻢ ﺍﻹﺣﺼﺎﺀ ﻭﺑﺤﻮﺙ ﺍﻟﻌﻤﻠﻴﺎﺕ
ﻃﺮﻕ ﺍﻟﺘﻨﺒﺆ ﺍﻹﺣﺼﺎﺋﻲ
) ا
ء اول (
ﺗﺄﻟﻴﻒ ﺩ .ﻋﺪﻧﺎﻥ ﻣﺎﺟﺪ ﻋﺒﺪﺍﻟﺮﺣﻤﻦ ﺑﺮﻱ
ﺃﺳﺘﺎﺫ ﺍﻹﺣﺼﺎﺀ ﻭﺑﺤﻮﺙ ﺍﻟﻌﻤﻠﻴﺎﺕ ﺍﻟﻤﺸﺎﺭﻙ
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
!%& '( ! رب ا
"
! وا
ة وا
م اف ا و%
ا
.)(* و%+
* و, و
.". &أ
.س12ر1
3(
ا45& ب6
78 59 ا:(';
;ب =ق ا3
دة او1 !
ا7@ هAه
@216;. م1Bف أ1
"C ذن اF. و7 دة إ
&ء ا1 & HI ;ب3
ا اAه
م1"
ان اJ; أ7 اA ه7 ;و3
9* ا3K. HI & ;! وH3K. *' %C@ و2 Cو
73C; &L آ;ب7N O"P ان وQ%. &; رعH3K.& و12 ر16;C 4'J;
وا
.S ; 9&ت وا1"!
رة ا1U 7N 4+ع و1P1!
ا43&'2 دV& W';2X
ع1P1&
ق إ6;2ع و1P1!
4
;ب ات او3
ا
ء اول & ا76Y2
دةL 4P2 ر4
"& اولS آ7;
واARIMA ;[ام !ذجF. 78 59 ا:(';
ا
SB6C ! آTime Series 4'&
;[ام ا
!; ت اF. 78 59 ا:(';
4!3%&و
V!L 7N و78 59 ا:(';
43;ر1O
ا42J;
ق ا6
"\ ا.
;ب إ3
ا, 7N
48 59 ا4&%
;[ام اF. 4ت ا
راX%
وا4]&^ اP1;. S!B ;ب3
اء اLأ
*وN V! . ء59
;") ) ا4+ رت1= W5 _&ا. 7 وهMinitab
.
! ن. ب6
ةN1;& 4&%
@ اAوه
ذن اF. ق6;2 ف1 "
ب ا
رات ا6
*L1&;ب و3
& ا7 ]
ا
ء ا
H21%;
ا4
و !ذج داIntervention Analysis H;
اH%C H]& VPا1!
ا
!;"دة4'&
و
!ذج ا
!; ت اTransfer Function Models
*OL1!
ا4'&
و
!ذج ا
!; ت اMultivariate Time Series Models
State !
و&^ آ4
%
`ء اN و
!ذجVector Time Series Models
Threshold Time %
و
!ذج اSpace Models and Kalman Filtering
:(';
ا7N OCJ(6C وGARCH و
!ذجARCH و
!ذجSeries Models
ت3(K
ق إ
ا6;' ! آFinantial Time Series Forecasting 7
!
ا
.78 59 ا:(';
ا7N O& وإ;[اNeural Networks 4( "
ا
4(;3!
اء اU9) و23
* اOL1
H!"
ا اA إ
ز ه7N 7'JN12 ا & ا ان1Lا وارAه
.;ب3
ا اA هH]&
ة إJb
ا4."
ا
2
S
;
9 ا43( Lا1;& ن13 1 ) وهW
= ي7
& ;ب3
ا اAن ه13
http://www.abarry.ws/books/statisticalForecast.pdf VB1!
ا7N
.N1!
وا ا
e
:!
ا
ي. !5
( اL&
ن.د
د1" f!
ا4"&L
هـ1422 "ةJ
ذو ا
م2002 2'2
3
ﺍﻟﻤﺤﺘﻮﻳﺎﺕ
&4&J
-1ا
H bاول 4&J& :و"Cر9...................................................................e2
أ&] 4ا
!; ت ا
&'9 ....................................................... 4 1-1
ا
Yض & درا 4و H%Cا
!; ت ا
&'9 ................................. 4 2-1
3-1ا
[16ات ا
!;[Aة
('ء
!1ذج 9 ................................................ :('C
9 "C 1-3-1ا
'!1ذج ...................................................................
(6C 2-3-1ا
'!1ذج 10 ..................................................................
o[KC 3-3-1وإ;(ر ا
'!1ذج 10 .....................................................
1C 4-3-1ا
;'(:ات 10 ....................................................................
5-3-1إ;[ام ا
;'(:ات وو VPا
Jارات 10 .........................................
"C 4-1ر e2و&(دئ أو
10 ................................................................... 4
10 7P& e2"C 1-4-1أو Cر r2ا
Iهة .............................................
10 e2"C 2-4-1ا
P%أو اsن .......................................................
10 e2"C 3-4-1أ6ء ا
;......................................................... (6
11 e2"C 4-4-1أ6ء ا
;'(............................................................ :
e2"C 5-4-1اJ;9ار 11 .................................................................
e2"C 6-4-1ا
` 4ا
(`ء 11 .........................................................
]& 7-4-1ل : 1ا
! 7Kا
"1Kا11 ................................................... 78
12 e2"C 8-4-1دا
4ا
; 2Yا
Aا.................................................... 7C
e2"C 9-4-1دا
4ا
;ا w.ا
Aا12 ................................................... 7C
]& 10-4-1ل : 2دا
4ا
;ا w.ا
Aا 4 `
7Cا
(`ء 12 .........................
e2"C 11-4-1دا
4ا
;ا w.ا
Aا 7Cا
13 ...................................... 78
]& 12-4-1ل :3دا
4ا
;ا w.ا
Aا 7Cا
4 `
78ا
(`ء 14 ..................
e2"C 13-4-1دا
4ا
;ا w.ا
Aا15 ......................................... 4'"
7C
e2"C 14-4-1دا
4ا
;ا w.ا
Aا 7Cا
16 .............................. 4'"
78
4
]& 15-4-1ل :4دا
4ا
;ا w.ا
Aا 7Cوا
;ا w.ا
Aا 7Cا
17 ......... 4'"
78
-2ا
H bا
]
!
:7ذج ا%
9ار ا
Aا-7Cا
!; w1ا
!;%ك ARMA
-3وإ;[ا& 7N OCا
;'(25 .........................................................................:
!
e2"Cذج ا%
9ار ا
Aا-7Cا
!; w1ا
!;%ك & ا
ر25 ....... (p,q) 4L 1-2
H& e2"Cا9زا 45ا
[25........................................................ 7b 2-2
H& e2"Cا9زا 45ا&&25 .................................................... 7 3-2
H& e2"Cا
;25 ............................................................... 2b 4-2
H& e2"Cا
; !25 ............................................................... V 5-2
أ&]26 ................................................................................... 4 6-2
!
o8ذج ا%
9ار ا
Aا-7Cا
!; w1ا
!;%ك 26 ........................... 7-2
1!
1-7-2ذج )26 ......................................................... ARMA(0,0
1!
2-7-2ذج )31 ................................................................... AR(1
1!
3-7-2ذج )31 ................................................................... AR(2
1!
4-7-2ذج )36 .................................................................... MA(1
1!
5-7-2ذج )39 ................................................................... MA(2
1!
6-7-2ذج )40 ........................................................... ARMA(1,1
1 7-7-2اص
!ذج )47 ............................................... ARMA(p,q
-3ا
H bا
]
!
:Qذج ا
!; ت ا
&' z 4ا
! ;Jة 58 ....................................
م اJ;9ار 7Nا
!;58 ............................................................ w1 1-3
م اJ;9ار 7Nا
;(59 ............................................................... 2 2-3
!ذج ا%
9ار ا
Aا-7Cا
;-7&3ا
!; w1ا
!;%ك & ا
ر62 ..... (p,d,q) 4L 3-3
1!
1-3-3ذج )62 ...................................................... ARIMA(1,1,0
1!
2-3-3ذج )62 ....................................................... ARIMA(0,1,1
1!
3-3-3ذج ا
! 7Kا
"1Kا
F. 78اف 63 ................................................
دا
4اوزان ) ψ (Bو!
H]!Cذج )63 .............................. ARMA(p,q 4- 3
ا&]
4ا
4اوزان
("\ ا
'!ذج 64 ..................................................... 5- 3
1-5-3دا
4اوزان
'!1ذج )64 .................................................... AR(1
2-5-3دا
4اوزان
'!1ذج )65 .................................................... MA(1
5
3-5-3دا
4اوزان
'!1ذج )65 .................................................... AR(2
4-5-3دا
4اوزان
'!1ذج )66 ................................................... MA(2
5-5-3دا
4اوزان
'!1ذج )66 ........................................... ARMA(1,1
6-5-3دا
4اوزان
'!1ذج )67 ................................................... ARI(1
7-5-3دا
4اوزان
'!1ذج ا
! 7Kا
"1Kا68 .................. ARIMA(1,0,1) 78
1 \".اص دا
4اوزان ) 69 .................................................... ψ (B 6- 3
-4ا
H bا
ا :V.ا
;'(:ات ذات &; V.& w1ا
[ {6اد
'!ذج )71 ....... ARMA(p,q
:2 42I
1-4أ6ء ا
;'(71 ......................................................................... :
41! & 2-4ا
!"&1ت 72 ................................................. Information Sets
:3 42I
3-4ا
!;'(| ذا &; V.& w1ا
[ {6اد
72 ............................................
B 4-4ة 72 ............................................................................................. 2
e2"C 5-4دا
4ا
;'(73 ................................................................................. :
6-4دوال ا
;'(!'
:ذج )73 ....................................................... ARIMA(p,d,q
1-6-4دا
4ا
;'(1!'
:ذج )73 ............................................................... AR(1
2-6-4ط ا!;9ار 73 ............................................................................
3-6-4دا
4ا
;'(1!'
:ذج )74 .............................................................. AR(2
4-6-4دا
4ا
;'(1!'
:ذج )74 ................................................. ARIMA(0,1,1
5-6-4دا
4ا
;'(1!'
:ذج )75 .............................................................. MA(1
6-6-4دا
4ا
;'(1!'
:ذج )75 .............................................................. MA(2
7-6-4دا
4ا
;'(1!'
:ذج )76 ..................................................... ARMA(1,1
5 7-4ود ا
;'(80 ......................................................................................... :
;N e2"C 1-7-4ة 4!J
:('Cا
! ;80 ..................................................... 4(J
]& 2-7-4ل 81 ........................................................................................
-5ا
H bا
[&~ )! C :و'.ء
Iم :('Cإ82 ............................................ 78 5
"C 1-5أو 2%Cا
'!1ذج 82 .......................................................................
S(]C 1-1-5ا
;(83 .............................................................................. 2
2-1-5إ;ر در 4Lا
;83 ................................................................ d 2b
84 ............................................................................... p,q 2%C 3-1-5
4-1-5إ )"& 4NPإ
اف 84 ...................................................................
6
2JCا
'!1ذج 85 .............................................................................. 2-5
4J2= 1-2-5ا
"وم 85 ........................................................................
2JC 2-2-5ا
"وم
("\ ا
'!ذج 86 .......................................................
1!'
1-2-2-5ذج )86 .............................................................. AR(1
1!'
2-2-2-5ذج )86 ......................................................... MA(1
1!'
3-2-2-5ذج )87 .......................................................... AR(2
1!'
4-2-2-5ذج )87 ......................................................... MA(2
1!'
5-2-2-5ذج )87 ................................................ ARMA(1,1
4J2= 3-2-5ا
!".ت ا
ا
89 .............................................. 4=K
2JC 4-2-5ات ا
!".ت ا
ا
\"(
4=Kا
'!ذج 89 .........................
!'
1-4-2-5ذج )89 .........................................................AR(1
!'
2-4-2-5ذج )90 ....................................................... MA(1
o[KCوإ;(ر ا
'!1ذج94 ......................................................... 3-5
o%N 1-3-5ا
(1ا94 .................................................................... 7B
1-1-3-5إ;(ر ا
!;1(
w1ا94 .................................................. 7B
2-1-3-5إ;(ر ا
"1Kا1(
48ا95 ................................................. 7B
3-1-3-5إ;(ر ا
;ا w.أو ا J;9ل
(1ا95 ................................... 7B
4-1-3-5إ;(ر =(" 4ا
(1ا96 .................................................... 7B
\". 2-3-5ا
!" 2اXى ;9ر
!1ذج &'96 ............................. W
1-2-3-5إ 48 5آ '
1و96 ........................................... ~3.
"& 2-2-3-5ر ا 9م ا
Aا96 ......................................... AIC 7C
3-3-5أ&] 4وX5ت درا96 .......................................................... 4
-6ا
H bا
دس!
:ذج ا%
9ار ا
Aا-7Cا
;-7&3ا
!; w1ا
!;%ك ا
!143 ...... 4!1
دوال ا
;ا w.ا
Aا 7Cوا
;ا w.ا
Aا 7Cا
\"(
78ا
'!ذج ا
!143 ..... 4!1 1-6
1!'
1-1-6ذج 144 ......................................... SARMA(0,1)(1,1)12
دوال ا
;ا w.ا
Aا \"(
7Cا
'!ذج ا
!145 .................................. 4!1 2- 6
145............................................. SARIMA(0,d,0)(0,D,1)s 1-2-6
145 ........................................... SARIMA(0,d,0)(1,D,1)s 2-2-6
145 ............................................ SARIMA(0,d,1)(0,D,1)s 3-2-6
7
146 ........................................... SARIMA(0,d,0)(1,D,1)s 4-2-6
146 ........................................... SARIMA(0,d,1)(1,D,0)s 5-2-6
146 ............................................ SARIMA(0,d,2)(0,D,1)s 6-2-6
دا
4ا
;ا w.ا
Aا 7Cا
1!'
78ذج ا
! 7!1ا
;`147 .................... 7b 3- 6
أ&] 4ا
!; ت ا
&' 4ا
!147 ....................................... 4!1 4- 6
إ;Jق دوال !
\"(
:('Cذج ا
!; ت ا
! 4!1ا
;`148 .......... 4b 5- 6
1-5-6دا
4ا
;'(1!'
:ذج 148 .................... SARIMA(0,0,0)(0,1,1)12
2-5-6دا
4ا
;'(1!'
:ذج 148 .................... SARIMA(0,1,1)(0,1,1)12
أ&] 4وX5ت درا ;!
4ت ا
&' 4ا
!149 .......................... 4!1 6- 6
ا
ء ا
"!:7
-7ا
H bا
:V.ورC 4Bر 7! W2ا
;'(1. :ا!
46ذج ا%
9ار
ا
Aا-7Cا
!; w1ا
!;%ك 178 .......................................................................
-8ا
H bا
]& ]& :ل H%Cا
(1ا 7Bو&" إ;ر
!1ذج &'193 .................. W
-9ا
H bا
; H%C :Vأو f3bCا
!; 4ا
&' 4إ
&آ(ت 202 .....................
-10ا
;! Oوا
;'(1. :ا 46ا
!; w1ا
!;%ك 225 ...........................................
1-10ا
w1ا
ري 225 .............................................................
-11ا
H bا
%دي :Kا
;! Oوا
;'(1. :ا 46ا
;! Oا 7ا
( 232 ............. w
-12ا
H bا
]
:K 7ا
;! Oوا
;'(1. :ا 46ا
;! Oا 7ا
!دوج 239 ...........
. 4J2= 1-12اون 239 ........................................................................
4J2= 2-12ه240 ........................................................................ S
1
3-12أ&]242 .................................................................................. 4
-13ا
H bا
]
:K Qا
;! Oا 7ا
] 7Uوا
;'(1. :ا4J2= 46
و
;ز
!; ت ا
! 4!1ا
!' 250 ..................................................... 4N
1-13ا
'!1ذج ا250 ............................................................... 7NP9
2-13ا
'!1ذج ا
;`242 .............................................................. 7b
]& 3-13ل
('ء
!1ذج 259 ........................................................... :('C
]& 4-13ل '(
,ء
!1ذج 267 ..................................................... :('C
& (1) %أ 4وإ.Lت ا(;9رات ا
275 ............................................... 4J.
ا
!ا358 ........................................................................................... VL
8
9
ﺍﻟﻔﺼﻞ ﺍﻷﻭﻝ
ور
:1
ا ا
Time Seriesه
#
$ا " ا ه ة !ه
ة ا
%ا
) #او
' ا)ن (
ا
' -ا,ت ا
:
" -1اbBل f'. )Oا
2ض .&12
-2د ا
51ات ا
! 4.16ا( & 1ا
;ج ".4'"& 4
) 5 -3ا
!("ت .& 4" & 2O
) 5 -4ا;
9ج ا
wb'
7&1ا
[م .43!!
.
وا
3ض
#درا 0و ./ا,ت ا
ه :
)ON -1و
!1K 4LAا 48ا
Iهة ا
!Kهة.
-2ا
;'( :ا
)Jا
! ;I
4(Jهة ا
"1Kا.48
-3ا
;I
. )3%هة ا
"1Kا 48إذا ا& 3ذ
.f
ا
H3Kا
;
4 ;!
7ز&'K& 4هة وه( 7رة ا;
9ج ا
J(
H=
. W%
7&1ة &
850
750
C1
650
550
Index 10 20 30 40 50 60 70
10
ا ;:ات ا9:ة ء 7ذج :4
إن إ 2د
!1ذج &' 4 ;& 4 (6'C Wز&'K& 4هة & (;"2ا
!Oم ا
"( 4وا
;;%C 7ج
ا
ا
& ]3ا
( Q%وا
[(ة1 .ف
;"ض \".ا
[16ات ا
"'(
4`2ء
!1ذج ر7P2
;'( 4 ;& :ز&':& 4
#ا ذج أو /ا ذج :Model Identification -1
وهAا ). );2ا
!; 4ا
&'132 Q5 Time Plot ! 2 !N 4ن ا59ا 7Uا 7JNXه 1ا
&
وا
أ ) 5 7ا
Iهة ا
!Kهة و& )Uإ;ر
!1ذج ر \". 7 2!;"& 7P2ا
!~2J
ا 48 59ا
; 1!
!Cذج ,و ا
[(ة ا
! ;!ة & ا
رات وا%.Xث.
;= ا ذج :Model Fitting -2
1!
^C ".ذج او اآ] آ'!1ذج &' e+1
Wا
!; 4ا
!Kهة
1Jم )
"& 2J;.هAا
ا
'!1ذج & ا
(
ت ا
!Kهة [;F.ام =ق ا
; 2Jا 78 59ا
[ ;!
. 4+ت ا
&'4
وهAا ا
'!1ذج ا
!^ A:2آ'!1ذج او
.J5X H2";
H.B 7
>:وإ?ر ا ذج :Model Diagnostics -3
إLاء إ;(رات 4 %bCأ6ء ا
;& 4N"!
Fitting Errors (6ى .6Cا
!Kهات
& Vا
)Jا
! & 4.1 %ا
'!1ذج ا
!^ و&ى PN 4%+ت ا
'!1ذج 4
5 7N .إ;Lز ا
'!1ذج
ا
!^
@AOا(;9رات
1Jم !;F.دة ا
* ا
'!1ذج ا
' 78Oو[; 2م
;:('C
1ات
)J
ا
! ; 4(Jوإ1"
Xد
[16ة اXو
;"
!1ذج .2L
ا4ات :Forecast Generation -4
5 )Uب أ6ء ا
;'(: ا
)Jا
! ; 4(Jو& [; 2م ا
'!1ذج ا
':('C
1;
78Oات
ا
!; 4ا
&' 4و&ا 4(Bه@A Forecast Errorsآ! ا; ت 2L )Bة &Kهة &
ا6ء 66[!. ! 2 & Nت ا
!ا Control Charts 4(Bوا
;1(J
VP1C 7ل {6 4( '.
&" إذا Cوز 4Cأ6ء ا
;'("2 :د ا
' 7N Iا
'!1ذج و"Cد ا
ورة & 1!
2%;. 2Lذج
&^ .,
إ :0ام ا4ات وو %Aا
ارات Implementation and Decision -5
JC :makingم ا
;'(:ات 7"
2JC Nا
Jار
' 7N Iإ;[ا&H3K
. O
ا
!'.W
11
: ر و
دئ او
&
. 4'&
ا4 ;!
ا1C 7;
ا48ا1K"
ا4!"
أو ا48ا1K"
هة اI
&
ف1
{Z t } 4= (. { اوZ t , t ∈ {⋯, −1,0,1,2,⋯}} او ا; را {⋯, Z −1 , Z 0 , Z1 , Z 2 ,⋯}
{z1 , z2 ,⋯, zn−1 , zn } &
. هةK!
ا4'&
ا4 ;!
و
:2
History ا!ه
ةE' او رA$ ' z1 , z2 ,⋯ , zn −1 " ا
4LA!'
ا4! 7N اL )O& r2وا
;ر
: 3
نG
او اA/ ' اzn ا
. هة اةK!
ا7وه
:4
) ; ه ا " اzˆt JK et = zt − zˆt , t = 1, 2,..., n I,$ '; =;أ?;ء ا
Residuals N0 ا
واO ا ذج( و أ#
L .M/7 ا " ا
.ذج1!'
ا2JC ". ة5 وا4"N دO H %
(6;
ء ا6 ان ا5 2و
مH3K. اوzn +1 , zn +2 , zn +3 ,... ز1&
. 4(J; !
هات اK!
& ف1 : !K,
مH3K. او zn (1) , zn ( 2 ) , zn ( 3) ,... &
. OCا:(';
& و zn + ℓ , ℓ ≥ 0
zn ( ℓ ) , ℓ ≥ 0
12
:5
en ( ℓ ) = zn +ℓ − zn ( ℓ ) , ℓ ≥ 0 I,$ '; 4أ?;ء ا
4JJ%
) اJ
هت ا1م ا
& وJC !ى آX ا1C ة5ا1
';_ اC :(';
ء ا6وأ
:6
S K إذاStationary {
ةz1 , z2 ,⋯ , zn −1 , zn } ل ان ا ا
ا ه ة
: ا
وط ا
1) E ( zt ) = constant = µ , ∀t
constant = γ 0 , ∀t , ∀s, t = s
2) cov ( zt , zs ) =
f ( s − t ) , ∀t , ∀s, t ≠ s
Building Blocks ب ا
('ء1= ة او5 O
13
اL 4!O& 4'& ز4 ;& ف
"ف1 نsا
Oف
ر1 7;
ا
'!ذج اV!
:7
White ءO اVO او اWhite Noise Series ءO اVO
ا
) ;$
ا
اW ا ه ات ا ا#
$
# { ه رةat } Noise Process
زتL
ات ا ا ا ) ن
و3 ا#
$
L7
ض اY7 7Kوا
#
ي وY[ \0 $ ( Independent, Identically Distributed (IID) $;
: أيσ 2 S$]
1) E ( at ) = 0, ∀t
σ 2 , ∀t , ∀s, t = s
2) cov ( at , a s ) =
0 , ∀t , ∀s , t ≠ s
at ~ WN ( 0,σ 2 )
$ L
و
13
:1ل-
: Random Walk
ا ا ا
:7
;
{ آZ t } 48ا1K 4! 7'( ف1
Z1 = a1
Z 2 = a1 + a2
⋮
Z t = a1 + a2 + ⋯ + at
أو
Z t = Z t −1 + at
Z t نN j &
' اe[
&م او اX ا7
اA:C 7;
ة ا16[
) ا5 1 هa j ا;(
ان1
أي
t &
' ا78ا1K 7& VB1& 7ه
4!
"
اق ا
!ل ا1 اe C 7;
ا اL 4&O
& ا
'!ذج ا4 ;!
او ا4!"
@ اA ه: !K,
ة؟J; & 4!"
اH وهt , s )B V!
cov ( Z t , Z s ) وE ( Z t ) L او: 2!C
:8
: و
ف آAutocovariance Function ا9
ا3دا ا
γ t ,s = cov ( Z t , Z s ) , ∀t , ∀s
= E ( Z t − µ )( Z s − µ ) , ∀t , ∀s
نca Z t + k أوZ t −k #$ وZ t #$ .MY ة ا
اY اb7 اk : اa وإذا
: I,$ '; ا9 ا3دا ا
γ k = cov ( Z t , Z t −k ) , k = 0, ±1, ±2,⋯
= E ( Z t − µ )( Z t −k − µ ) , k = 0, ±1, ±2,⋯
!8 دا7
]
اe2";
ف
;[م ا1 : !K,
14
:9
: و
ف آAutocorrelation Function (ACF) ا9\ ا$دا ا
ا
γk
ρk = , k = 0, ±1, ±2,⋯
γ0
:4
;
اص ا1[
اO
و
1. ρ 0 = 1
2. ρ − k = ρ k
3. ρk ≤ 1
:2 ل-
ا
(`ء4 `
ا4!"
7CاA
اw. ا
;ا4
ن داs; اK
ف1
:7 ا
(`ء ه4 `
ا4!"
7CاA
ا2Y;
ا4
دا
σ 2 , k = 0
γ k = cov ( at , at −k ) =
0, k ≠0
:7CاA
اw. ا
;ا4
داO'& و7 e2";
& اf
وذ
γ k 1 , k = 0
ρk = =
γ 0 0 , k ≠ 0
:7
;
اH3K
اO
و
Autocorrelation function of White Noise
1.0
Autocorr
0.5
0.0
0 1 2 3 4 5 6 7 8 9
Lag
15
:10
Partial Autocorrelation Function (PACF) Vا ا9\ ا$دا ا
ا
ات3 ا#
f\ ا$]
ا
اg S إزا$ Z t −k وZ t #$ \$و;
ار ا
ا
قi K وأφkk
$ k : اL
وL$ I ا اZ t −1 , Z t −2 ,..., Z t −k +1
:.- اa φkk V ار ا/7j ا.
بK مL$K
Z t = φk 1Z t −1 + φk 2 Z t −2 + ⋯ + φkk Z t −k + at
: φ11 ب5
Z t = φ11Z t −1 + at
VB1;
اA وأZ t −1 ـ. 4B "
ا7N= `ب.
E ( Z t −1 Z t ) = φ11 E ( Z t −1 Z t −1 ) + E ( Z t −1at )
أي
γ 1 = φ11γ 0
( J5X (' ! آE ( Z t −k at ) = 0, k = 1, 2,... مH3K. ) E ( Z t −1at ) = 0 Q5
γ 0 7 4! J
.و
φ11 = ρ1
16
:11
: آVا ا9\ ا$ م
ف دا ا
ا.) $
1, k =0
ρ1 , k =1
1 ρ1 ⋯ ρ k −2 ρ1
ρ1 1 ⋯ ρ k −3 ρ2
⋮ ⋮ ⋯ ⋮ ⋮
φkk = ρ k −1 ρ k −2 ⋯ ρ1 ρ k
, k = 2,3,...
1 ρ1 ⋯ ρ k −2 ρ k −1
ρ1 1 ⋯ ρ k −3 ρ k −2
⋮ ⋮ ⋯ ⋮ ⋮
ρ k −1 ρ k −2 ⋯ ρ1 1
4N1b & دة%& 7
& اC Q5
4
ب دا%
, e2"C 76"
ف1 اAO
(ة و3
اk )J
;[ام9 اW"+ .
اe2";
ا
:2ار3C 78
ا7CاA
اw.ا
;ا
: ب11
تI, ا#
)
ارφkk N/
φ00 = 1, by definition
φ11 = ρ1
k −1
ρ k − ∑φk −1, j ρ k − j
j =1
φkk = k −1
, k = 2,3,...
1 − ∑φk −1, j ρ j
j =1
JK
φkj = φk −1, j − φkkφk −1,k −1 , j = 1, 2,..., k − 1
17
: φ22 ب5
: ب11 e2"C &
ρ 2 − φ11 ρ1 ρ 2 − ρ12
φ22 = =
1 − φ11 ρ1 1 − ρ12
. φ11 = ρ1 نf
وذ
:3 ل-
: ا
(`ء4 `
ا4!"
78
ا7CاA
اw. ا
;ا4
ن داs; اK
ف1
ب11 e2"C &
φ00 = 1, by definition
φ11 = ρ1 = 0
.
ا1 & &]لf
وذ
φkk ب11 e2"C 7N \21";
.و
φ22 = φ33 = ⋯ = 0
:اA3وه
1, k = 0
φkk =
0, k ≠ 0
:7
;
اH3K
اO
و
18
Partial Autocorrelation function of White Noise
1.0
PACF
0.5
0.0
0 1 2 3 4 5 6 7 8 9
Lag
ا
(`ء4 `
ا4!"
78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
& داH أن آ5X : !K,
او46. ا
!;اz 48ا1K"
ات اY;!
اV!L 4+ @A وه. اولe[;
& اb
وي اC
.f
A
7CاA
اw. ا
;ا4
;[م داC 78ا1K Y;!
هةK& )B . w.;(ر م ا
;ا9 .4J; !
ا
: 12
Sample Autocorrelation Function SACF ا9\ ا$دا ا
ا
: I,$ '; وrk , k = 0,1, 2,...
$ L
وz1 , z2 ,⋯ , zn −1 , zn ز
ه ة
n −k
∑( z t − z )( zt +k − z )
rk = t =1
n
, k = 0,1,2,...
∑( z −z)
2
t
t =1
1 n
z= ∑ zt JK
n t =1
رJ&ُ O ! ا. وρˆ k = rk , k = 0,1, 2,... أي7CاA
اw. ا
;ا4
اEstimator رJ&ُ 7وه
:4
;
ا4'"
اص ا1[
اO
نFN اAO
ى وX 4' & 8ا1K Y;C إذا7ON
نFN ρ k = 0, k > q S إذا آ-1
1 q
V ( rk ) ≅ 1 + 2∑ ρ k2 , k > q
n k =1
19
1
V ( rk ) ≅ , k > 0 نFN ρ k = 0, k > 0 &' 4+[
ا4
%
ا7Nو
n
مJ
اV6;
7
;
. و7"(= V2ز1C (2JC O
ن132 rk نFN ρ k = 0 (ة و3
اn )J
-2
:7
;
;(ر اX.
H 0 : ρk = 0
H1 : ρ k ≠ 0
:48 59;[ام اF. f
وذ
rk
− 12
= n rk
n
n rk > 1.96 S
إذا آH 0 \NC وα = 0.05 421'"& ى1; & ' f
وذ
corr ( rk , rk − s ) ≅ 0, s ≠ 0 نFN H 0 : ρ k = 0, ∀k 4Pb
اS%C -3
:7
;
آ4'"
7CاA
اw. ا
;ا4
'ت
ا2(;
ر اJCُ -4
1 q
Vˆ ( rk ) ≅ 1 + 2∑ rk2 , k > q
n k =1
:13
Sample Partial Autocorrelation Function Vا ا9\ ا$دا ا
ا
$ L
و z1 , z2 ,⋯ , zn −1 , zn
ه ة
ز SPACF
: I,$ '; rkk , k = 0,1, 2,...
20
1, k =0
r, k =1
1
1 r1 ⋯ rk −2 r1
r1 1 ⋯ rk −3 r2
⋮ ⋮ ⋯ ⋮ ⋮
rkk = rk −1 rk −2 ⋯ r1 rk
1 , k = 2,3,...
r1 ⋯ rk −2 rk −1
r1 1 ⋯ rk −3 rk −2
⋮ ⋮ ⋯ ⋮ ⋮
rk −1 rk −2 ⋯ r1 1
:2ار3C 4'"
78
ا7CاA
اw. ا
;ا4
ب دا%
و
: ب13
تI, ا#
)
ارrkk N/
r00 = 1, by definition
r11 = r1
k −1
rk − ∑ rk −1, j rk − j
j =1
rkk = k −1
, k = 2,3,...
1 − ∑ rk −1, j rj
j =1
JK
rkj = rk −1, j − rkk rk −1,k −1 , j = 1, 2,..., k − 1
أي 4'"
78
ا 7CاA
ا w.ا
;ا 4
ا Estimator رJ& `2ا 7وه
نFN اAO
ى وX 4' & 8ا1K Y;C إذا7ON
رJ&ُ O
! ا. وφˆkk = rkk , k = 0,1, 2,...
:4
;
ا4'"
اص ا1[
اO
21
1
V ( rkk ) ≅ , k > 0 -1
n
;(رX. مJ
اV6;
7
;
. و7"(= V2ز1C (2JC O
ن132 rkk نFN (ة3
اn )J
-2
:7
;
ا
H 0 : φkk = 0
H1 : φkk ≠ 0
:48 59;[ام اF. f
وذ
rkk
− 12
= n rkk
n
n rkk > 1.96 S
إذا آH 0 \NC وα = 0.05 421'"& ى1; & ' f
وذ
corr (φkk ,φk −s ,k − s ) ≅ 0, s ≠ 0 نFN H 0 : φkk = 0, ∀k 4Pb
اS%C -3
:7
;
آ4'"
7CاA
اw. ا
;ا4
'ت
ا2(;
ّر اJCُ -4
1
Vˆ ( rkk ) ≅ , k > 0
n
:4 ل-
:&12 "& _;'& 7 W6
اH]!C 4
;
ا
( ت ا
158 222 248 216 226 239 206 178 169
:!O! وار4'"
78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;اW 5أ
1 n 1
z = ∑ zt = (158 + 222 + ⋯ + 169 ) = 206.89 w1;!
اW %
:Xاو
n t =1 9
4B "
& ا7CاA
اw. ا
;اW %
:
U
n −k
∑( z t − z )( zt +k − z )
rk = t =1
n
, k = 0,1,2,...
∑( z −z)
2
t
t =1
22
(158 × 222 + 222 × 248 + ⋯ + 178 × 169 )
r1 = = 0.265116
(158 × 158 + 222 × 222 + ⋯ + 169 × 169 )
(158 × 248 + 222 × 216 + ⋯ + 206 × 169 )
r2 = = -0.212
(158 × 158 + 222 × 222 + ⋯ + 169 × 169 )
(158 × 216 + 222 × 226 + ⋯ + 239 × 169 )
r3 = = −0.076
(158 × 158 + 222 × 222 + ⋯ + 169 × 169 )
r8 = 0.230, r7 = 0.104, r6 = −0.242, r5 = −0.387, r4 = −0.183 اA3وه
& 'ت2(;
اW %
:]
U
1 q
Vˆ ( rk ) ≅ 1 + 2∑ rk2 , k > q
n k =1
1
Vˆ ( r1 ) ≅
9
n
1
9
(
Vˆ ( r2 ) ≅ (1 + 2r12 ) = 1 + 2 ( 0.265) = 0.1267
1 2
)
1
n
( 1
9
) ( 2
(
Vˆ ( r3 ) ≅ 1 + 2 ( r12 + r22 ) = 1 + 2 ( 0.265) + ( −0.212 )
2
)) = 0.1367
Vˆ ( r4 ) ≅ 0.138 Vˆ ( r5 ) ≅ 0.1454 Vˆ ( r6 ) ≅ 0.1787
…r
ا
:4'"
78
ا7CاA
اw. ا
;اW %
:".را
r00 = 1, by definition
r11 = r1 = 0.265
42ار3;
ت اB "
ت & ا6. ا
;ا7B. W %
)U
k −1
rk − ∑ rk −1, j rk − j
j =1
rkk = k −1
, k = 2,3,...
1 − ∑ rk −1, j rj
j =1
Q5
rkj = rk −1, j − rkk rk −1,k −1 , j = 1, 2,..., k − 1
23
1
r2 − ∑ r1, j r2− j
j =1 r2 − r11r1 ( −0.212 ) − ( 0.265)( 0.265) −0.282225
r22 = = = =
1
1 − r11r1 1 − ( 0.265)( 0.265) 0.929775
1 − ∑ r1, j rj
j =1
= −0.30354
& W %C وr21
;ج ا%
r33 ب%
r21 = r11 − r22 r11 = 0.265 − ( −0.303)( 0.265) = 0.345295
2
r3 − ∑ r2, j r3− j
j =1 r3 − ( r21r2 + r22 r1 )
r33 = =
k −1
1 − ( r21r1 + r22 r2 )
1 − ∑ r2, j rj
j =1
( −0.076 ) − ( ( 0.345)( −0.212 ) + ( −0.303)( 0.265) )
=
1 − ( ( 0.345)( 0.265) + ( −0.303)( −0.212 ) )
= 0.092
4'"
48
ت ا6. ا
;ا7B. W % اA3وه
r88 = 0.042, r77 = 0.013, r66 = −0.207, r55 = −0.294, r44 = −0.298
1 = 0.1111 (2JC ويC 'ت2(;
!" اL O
و
9
4'"
78
ا7CاA
اw. وا
;ا7CاA
اw. ر) دوال ا
;ا: &
Autocorrelation Function for Demand
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8
Lag Corr T LBQ Lag Corr T LBQ
1 0.27 0.80 0.87 8 0.23 0.52 13.66
2 -0.21 -0.59 1.50
3 -0.08 -0.21 1.60
4 -0.18 -0.49 2.26
5 -0.39 -1.01 5.96
6 -0.24 -0.57 7.89
7 0.10 0.24 8.43
24
Partial Autocorrelation Function for Demand
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8
Lag PAC T Lag PAC T
1 0.27 0.80 8 0.04 0.13
2 -0.30 -0.91
3 0.09 0.27
4 -0.30 -0.89
5 -0.29 -0.88
6 -0.21 -0.62
7 0.01 0.04
25
ﺍﻟﻔﺼﻞ ﺍﻟﺜﺎﻧﻲ
Autoregressive-Moving Average
ك/\ ا0 ا_ا9 ار ا/7jذج ا7
:4 اa L
ا:0 وإModels
ك%;!
اw1;!
_ا7CاA
ار ا%
9
!ذج اO 62 7;
ا
'!ذج ا & آ(ة48 ه'ك
e;[& 7N ]ة3
ث ا%. اS;(U ا7;
واAutoregressive-Moving Average Models
.:(';
ا7N 42J;
ق ا6
ا7 H8O
اOB1bC 7 4J(6;
ا2ا
!د
:14
ARMA ( p, q ) b
و ( p, q ) n ا ر#
ك/\ ا0 ا_ا9 ار ا/7j ذج ا7
:.) اN) {z1 , z2 ,… , zn −1 , zn } ز
ه ة
zt = δ + φ1 zt −1 + φ2 zt −2 + ⋯ + φ p zt − p + at − θ1at −1 − θ 2 at −2 − ⋯ − θ q at −q
ا ي.- S$] "
−∞ < δ < ∞ ء وO$ VA
at ~ WN ( 0,σ 2 ) JK
وAutoregressive Parameters ا9 ار ا/7j ه
" اφ1 ,φ2 ,… ,φ p و
Moving Average Operators ك/\ ا0 ه
" اθ1 ,θ 2 ,… ,θ q
O"& H&";
اHO 2 73
@ ا
'!ذجA هw (;
Operators Algebra ( ا
"!ل. ";
ف1
اص: اb وB b
وBackshift Operator Y: اKزاj ا.
:15
: ا
1 − Bzt = zt −1
2 − B m zt = B m −1 ( Bzt ) = B m−2 ( B ( Bzt ) ) = ⋯ = zt −m
3 − Bc = c, c is a constant
:7 هJ5X O
;ج ا% !ل ايL1C 7b[
ا45زا9 اH& 7
ا4NP9.
26
: ب15
: و
ف آF b
وForewardshift Operator
q اKزاj ا.
-1
F = B −1
: ∇ و
ف آb
وDifference Operator =
Y ا.
-2
∇ = (1 − B )
: و ف آS b
وSum Operator %V ا.
-4
S = ∇ −1 = (1 − B )
−1
7 *(;3 و ( p, q ) 4Lك & ا
ر%;!
اw1;!
_ا7CاA
ار ا% 9ذج ا1! 7
د ا1" نsا
:H3K
ا
zt − φ1 zt −1 − φ2 zt −2 − ⋯ − φ p zt − p = δ + at − θ1at −1 − θ 2 at −2 − ⋯ − θ q at −q
zt − φ1 Bzt − φ2 B 2 zt − ⋯ − φ p B p zt = δ + at − θ1 Bat − θ 2 B 2 at − ⋯ − θ q B q at
(1 − φ B − φ B
1 2
2
− ⋯ − φ p B p ) zt = δ + (1 − θ1 B − θ 2 B 2 − ⋯ − θ q B q ) at
أو
φ p ( B ) z t = δ + θ q ( B ) at
Autoregressive 7CاA
ار ا%
9 اH& 1 هφ p ( B ) = 1 − φ1 B − φ2 B 2 − ⋯ − φ p B p Q5
ك%;!
اw1;!
اH& 1 هθ q ( B ) = 1 − θ1 B − θ 2 B 2 − ⋯ − θ q B q وOperator
Moving Average Operator
: -
أ
W;32 وARMA ( 0,0 ) *
&2 وConstant Mean Model S.]
اw1;!
ذج ا1!
-1
:H3K
ا7
φ 0 ( B ) z t = δ + θ 0 ( B ) at
او
(1) zt = δ + (1) at
zt = δ + at , at ~ WN ( 0,σ 2 )
27
:H3K
ا7 1 وهARMA (1,0 ) ≡ AR (1) 7
وX ا4L & ا
ر7CاA
ار ا%
9ذج ا1!
-2
φ1 ( B ) zt = δ + θ 0 ( B ) at
(1 − φ1B ) zt = δ + at
zt = δ + φ1 zt −1 + at , at ~ WN ( 0,σ 2 )
:H3K
ا7 1 وهARMA ( 0,1) ≡ MA (1) 7
وX ا4Lك & ا
ر%;!
اw1;!
ذج ا1!
-3
φ0 ( B ) zt = δ + θ1 ( B ) at
zt = δ + (1 − θ1B ) at
zt = δ + at − θ1at −1 , at ~ WN ( 0,σ 2 )
:H3K
ا7 1 وهARMA ( 2,0 ) ≡ AR ( 2 ) 4
]
ا4L & ا
ر7CاA
ار ا%
9ذج ا1!
-4
φ 2 ( B ) zt = δ + θ 0 ( B ) a t
(1 − φ B − φ B ) z
1 2
2
t = δ + at
zt = δ + φ1 zt −1 + φ2 zt −2 + at , at ~ WN ( 0,σ 2 )
7 *(;3 وARMA (1,1) (1و1) 4Lك & ا
ر%;!
اw1;!
_ا7CاA
ار ا% 9ذج ا1! -5
:H3K
ا
φ1 ( B ) zt = δ + θ1 ( B ) at
(1 − φ1B ) zt = δ + (1 − θ1B ) at
zt = δ + φ1 zt −1 + at − θ1at −1 , at ~ WN ( 0,σ 2 )
28
:
ك/\ ا0 ا_ا9 ار ا/7jذج ا7 > M?
ك%;!
اw1;!
_ا7CاA
ار ا%
9!
!ذج اC 7;
ا48 59 اo8 [
ف
رس ا1
ذج1!
2%C
;" اوf
هة وذK& 4' & @ ا
'!ذجA ه5 ا7 ا
;"ف4b آ4N"&و
.هاتK!
اe 2 W'&
:ARMA(0,0)S$-\ ا0 ذج ا7 :rأو
H3K
ا7 W;32و
φ 0 ( B ) z t = δ + θ 0 ( B ) at
او
zt = δ + at , at ~ WN ( 0,σ 2 )
w. ا
;ا7;
( وداw1;!
)اVB1;
د ا2F. f
ذج وذ1!'
ا اAO
48 59اص ا1[
; اK
ف1
:7
;
آ78
ا7CاA
اw. وا
;ا7CاA
ا
E ( z t ) = δ + E ( at )
=δ
at ~ WN ( 0,σ 2 ) نf
وذ
δ = µ ن132 7
;
. وµ = E ( zt ) أيµ &
. E ( zt ) 4 ;!
اw1;!
&
ف1
:ذج1!'
اW;32و
z t − µ = at
7N 4J.
ا4
ا
!"د7N=
`ب78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
ق داJ;9
أيVB1;
اA{
وzt −k − µ
E ( zt −k − µ )( zt − µ ) = E ( zt −k − µ ) at
إذا8 e2"C & E ( zt −k − µ )( zt − µ ) = γ k 3
و
γ k = E ( zt −k − µ ) at , k = 0, ±1, ±2,⋯
:2ار3C 4B "
@ اA هH% و
k = 0 : γ 0 = E ( zt − µ ) at
29
أيVB1;
اA{
وat 7N zt − µ = at 7N= !
`ب2ف ا6
د ا29
E ( zt − µ ) at = E ( at at ) = σ 2
إذاat ~ WN ( 0,σ 2 ) نf
وذ
k = 0 : γ 0 = E ( zt − µ ) at = σ 2
k = 1: γ 1 = E ( zt −1 − µ ) at = 0
نf
وذ
zt −1 − µ = at −1
E ( zt −1 − µ ) at = E ( at −1at ) = 0
نFN 4JJ%
ا7N
zt −k − µ = at −k , k = 1,2,…
E ( zt −k − µ ) at = E ( at −k at ) = 0, k = 1, 2,…
أي
:1 ةI
σ 2 , k = 0
E ( zt −k − µ ) at = E ( at −k at ) =
0, k = 1,2,..
أي
γ0 =σ 2
γ k = 0, k = ±1, ±2,…
:7
داH3 7 VP1Cو
σ 2 , k = 0
γk =
0, k ≠ 0
γ 0 = σ 2 7 4! J
.و
30
γ k 1, k = 0
ρk = =
γ 0 0, k ≠ 0
:7
;
اH3K
اO
و
Autocorrelation function of Constant Mean Model
1.0
Autocorr
0.5
0.0
0 1 2 3 4 5 6 7 8 9
Lag
11 e2";
& ا78
ا7CاA
اw. ا
;ا4
ن داs; اK
φ00 = 1, by definition
φ11 = ρ1 , by definition
φ11 = 0
1 ρ1 1 0
ρ1 ρ2 0 0
φ22 = = =0
1 ρ1 1 0
ρ1 1 0 1
31
1 ρ1 ρ1 1 0 0
ρ1 1 ρ2 0 1 0
ρ2 ρ1 ρ3 0 0 0
φ33 = = =0
1 ρ1 ρ2 1 0 0
ρ1 1 ρ1 0 1 0
ρ2 ρ1 1 0 0 1
⋮
1 ρ1 ⋯ ρ1 1 0 ⋯ 0
ρ1 1 ⋯ ρ2 0 1 ⋯ 0
⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
ρ k −1 ρ k −2 ⋯ ρk 0 0 ⋯ 0 0
φkk = = = = 0, k = 2,3,⋯
1 ρ1 ⋯ ρ k −1 1 0 ⋯ 0 1
ρ1 1 ⋯ ρ k −2 0 1 ⋯ 0
⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
ρ k −1 ρ k −2 ⋯ 1 0 0 ⋯ 1
:7
داH3 7 VP1Cو
1, k = 0
φkk =
0, k ≠ 0
:7
;
اH3K
اO
و
Partial Autocorrelation function of Constant Mean Model
1.0
PACF
0.5
0.0
0 1 2 3 4 5 6 7 8 9
Lag
z w1;& *
ان7N X ا
(`ء ا4 `
ذج ا1!
;قb2X S.]
اw1;!
ذج ا1!
: !K,
يb+
32
ARMA(1,0) = AR(1) وr اn ا ر#
ا9 ار ا/7j ذج ا7 :7]
:H3K
ا7 1وه
φ1 ( B ) zt = δ + θ 0 ( B ) at
(1 − φ1B ) zt = δ + at
zt = δ + φ1 zt −1 + at , at ~ WN ( 0,σ 2 )
:78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
( وداw1;!
)اVB1;
اL1
ف1 .
ذج ا1!'
آ
(1 − φ1 B ) zt = δ + at
δ
+ (1 − φ1 B ) at
−1
zt =
(1 − φ1 )
δ
E ( zt ) = + E (1 − φ1 B ) at
−1
(1 − φ1 )
1! ه2ف ا6
ا7N 7 ]
ا%
ا
∞ j
E (1 − φ1 B ) at = E ∑φ1 B j at
−1
j =0
∞
∑φ
j =0
1
j
B j < ∞ 48O
ا4 ;!
ن ا13C انW 2 78O
ع ا1! !
ا7 VB1;
دل ا9
Y;& دورW"2 نs اB H&"
إذا ا;( اf
وذφ1 < 1 S إذا آJ%;2 f
وذ4.رJ;&
ان.X 4JJ%
ا7N B = 1 سJ
و
* اB = a + ib H3K
* اComplex Variable W&آ
أيB > 1 ة أي51
ة ا8( رج دا1 − φ1 B ) = 0 رb+ور او اL ن13C انW6;
1 − φ1 B = 0
1
B=
φ1
1
B >1⇒ > 1 ⇒ φ1 < 1
φ1
4B "
ا7
د ا1"
.ارJ;9 ط ا1ا هAوه
33
∞ j j
E (1 − φ1 B ) at = E ∑φ1 B at
−1
j =0
∞ j
= ∑φ1 B j E ( at )
j =0
=0, ∀t
ن132و
δ
E ( zt ) =
(1 − φ1 )
او
δ
µ=
(1 − φ1 )
∴δ = µ (1 − φ1 )
ذج1!'
ا4Y+ 7N δ \21";
.و
zt = δ + φ1 zt −1 + at
= µ (1 − φ1 ) + φ1 zt −1 + at
= µ + φ1 ( zt −1 − µ ) + at
( zt − µ ) − φ1 ( zt −1 − µ ) = at
أيVB1;
اA{
وzt −k − µ 7N 4J.
ا4
ا
!"د7N=
`ب
E ( zt −k − µ )( zt − µ ) − φ1 E ( zt −k − µ )( zt −1 − µ ) = E ( zt −k − µ ) at , k = 0, ±1, ±2,⋯
أي
γ k − φ1γ k −1 = E ( zt −k − µ ) at , k = 0, ±1, ±2,⋯
:72 ! آ2ار3C 4B "
@ اA هH%C و8 e2"C & f
وذ
k = 0 : γ 0 − φ1γ 1 = E ( zt − µ ) at
:7
;
. م1J
!2ف ا6
د ا29
34
E at ( zt − µ ) − φ1 E at ( zt −1 − µ ) = E ( at at )
E at ( zt − µ ) − φ1 × ( 0 ) = σ 2
∴ E at ( zt − µ ) = σ 2
إذا
γ 0 − φ1γ 1 = σ 2
k = 1: γ 1 − φ1γ 0 = E ( zt −1 − µ ) at = 0
4JJ%
ا7N
γ k − φ1γ k −1 = 0, k = 1, 2,⋯
γ 0 7 اة4
ا
!"د4! J.
ρ k − φ1 ρ k −1 = 0, k = 1,2,⋯
أو
ρ k = φ1 ρ k −1 , k = 1, 2,⋯
:نFN ρ 0 = 1 ! ان.و
ρ1 = φ1 ρ 0 = φ1
ρ 2 = φ1 ρ1 = φ12
⋮
ρ k = φ1k
4
داH3K. أو
ρ k = φ1k , k = 0, ±1, ±2,⋯
أيρ k & WL1!
اK
ا+ن وs & اI'
ف1 ρ − k = ρ k , ∀k نf
وذ
ρ k = φ1k , k = 0,1, 2,⋯
:7
;
اH3K
اO
4
@ ا
اAه
φ1 > 0 ن13C &' -1
35
Autocorrelation function of AR(1) Model
0.5
0.4
0.3
ACF
0.2
0.1
0.0
0 1 2 3 4 5 6 7 8 9 10
Lag
φ1 < 0 ن13C &' -2
Autocorrelation function of AR(1) Model
0.3
0.2
0.1
0.0
ACF
-0.1
-0.2
-0.3
-0.4
-0.5
0 1 2 3 4 5 6 7 8 9 10
Lag
78
ا7CاA
اw. ا
;ا4
ن داs; اK
11 e2"C &
36
φ00 = 1, by definition
φ11 = ρ1 = φ1 , by definition
1 ρ1 1 φ1
ρ1 ρ 2 φ1 φ12 0
φ22 = = = =0
1 ρ1 1 φ1 1 − φ12
ρ1 1 φ1 1
⋮
1 ρ1 ⋯ ρ1 1 φ1 ⋯ φ1
ρ1 1 ⋯ ρ2 φ1 1 ⋯ φ12
⋮ ⋮ ⋯ ⋮ ⋮ ⋮ ⋯ ⋮
ρ k −1 ρ k −2 ⋯ ρk φ1k −1 φ1k −2 ⋯ φ1k 0
φkk = = =
1 ρ1 ⋯ ρ k −1 1 φ1 ⋯ φ 1
k
>0
ρ1 1 ⋯ ρ k −2 φ1 1 ⋯ φ 1
k −1
⋮ ⋮ ⋯ ⋮ ⋮ ⋮ ⋯ ⋮
ρ k −1 ρ k −2 ⋯ 1 φ1k −1 φ1k −2 ⋯ 1
W;3 وφ1 7N .د اول &`و1&"
وي ا2 د ا1&"
ا ن اb+ ويC w (
دة ا%&
:7
ا
اH3K
ا7 78
ا7CاA
اw. ا
;ا4
دا
1, k = 0
φkk = φ1 , k = 1
0, k ≥ 2
:7
;
اH3K
اO
و
φ1 > 0 ن13C &' -1
Partial Autocorrelation function of AR(1) Model
0.5
0.4
PACF
0.3
0.2
0.1
0.0
0 1 2 3 4 5 6 7 8 9 10
Lag
37
φ1 < 0 ن13C &' -2
Partial Autocorrelation function of AR(1) Model
0.0
-0.1
PACF -0.2
-0.3
-0.4
-0.5
0 1 2 3 4 5 6 7 8 9 10
Lag
.4
(
ل ا3 ا7N φ00 = 1 اوρ 0 = 1 & ) أيCX !8 دا:4I5 &
: ا ذجI
t )B V!
S.U1 وهE ( zt ) = δ (1 − φ1 ) نFN (ارJ;9 )ط اφ1 < 1 ن13C &' -1
t &
ا7 !;"CX وwJN k e[;
4
دا7CاA
اw. ا
;ا4
دا-2
&[;C وφ1 > 0 ن13C &' ρ1 & ;اءا. إ5 @ واC إ7N ;[& اC 7CاA
اw. ا
;ا4
دا-3
φ1 < 0 ن13C &' 4(
وا4(L1!
) اJ
ا. ا &;ددة
ن132 ( وφ00 7
اI'
م اV& ) 42b+ z ة5 وا4!B O
78
ا7CاA
اw. ا
;ا4
دا-4
φ1 وي2 ارهJ& وφ1 إرةW 5 O هCإ
: ARMA(2,0) = AR(2) 7- اn ا ر#
ا9 ار ا/7j ذج ا7 :-]
:H3K
ا7 W;32و
φ 2 ( B ) z t = δ + θ 0 ( B ) at
(1 − φ B + φ B ) z
1 2
2
t = δ + at
zt = δ + φ1 zt −1 + φ2 zt −2 + at , at ∼ WN ( 0, σ 2 )
:78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
وداw1;!
اL1
.
آ
38
(1 − φ B − φ B ) z
1 2
2
t = δ + at
δ
+ (1 − φ1 B − φ2 B 2 ) at
−1
zt =
(1 − φ1 − φ2 )
δ
+ E (1 − φ1 B − φ2 B 2 ) at
−1
E ( zt ) =
(1 − φ1 − φ2 )
∞
VB1;
اH
73
وE ∑ψ j at − j H3K
ا78O
X ع1! & !2ف ا6
ا7N 7
]
ا%
ا
j =0
∞
إذاJ%;2 اA وهV.!
اw1;!
ا7N 4.رJ;& ∑ψ
j =0
a
j t− j ن13C ان.X 78O
اV! ;
اHدا
∞
:4
;
وط اK
ا7CاA
ار ا%
9 &"
) اSJJ5 إذاJ%;2 اAوه ∑ψ
j =0
2
j < ∞ إذا آنwJNو
φ2 − φ1 < 1
φ2 + φ1 < 1
−1 < φ2 < 1
رb+ور او أL ن1آ & `2';_ اC وطK
@ اA
ار ) ه0j
وط ا$ 7! C 7;
وا
نFN ارJ;9 وط اSJJ%C إذا. ( ة51
ة ا8( رج دا1 − φ1 B − φ2 B 2 ) = 0
E (1 − φ1 B − φ2 B 2 ) at = (1 − φ1 B − φ2 B 2 ) E ( at ) = 0, ∀t
−1 −1
ن132و
δ
µ = E ( zt ) =
(1 − φ1 − φ2 )
δ = (1 − φ1 − φ2 ) µ
ذج1!'
ا4Y+ 7N δ \21";
. و
zt = (1 − φ1 − φ2 ) µ + φ1 zt −1 + φ2 zt −2 + at
= µ + φ1 ( zt −1 − µ ) + φ2 ( zt −2 − µ ) + at
( zt − µ ) − φ1 ( zt −1 − µ ) − φ2 ( zt −2 − µ ) = at
:
VB1;
اA{
وzt −k − µ 7N 4J.
ا4
`ب ا
!"د
39
E ( zt − µ )( zt −k − µ ) − φ1 ( zt −1 − µ )( zt −k − µ ) − φ2 ( zt −2 − µ )( zt −k − µ )
= E at ( zt −k − µ ) , k = 0, ±1, ±2,...
أي
E ( zt − µ )( zt −k − µ ) − φ1 E ( zt −1 − µ )( zt −k − µ ) − φ2 E ( zt −2 − µ )( zt −k − µ )
= E at ( zt −k − µ ) , k = 0, ±1, ±2,...
أو
γ k − φ1γ k −1 − φ2γ k −2 = E at ( zt −k − µ ) , k = 0, ±1, ±2,...
:72 ! آ2ار3C 4B "
@ اA هH%
نs ا8 e2"C & f
وذ
k = 0 : γ 0 − φ1γ −1 − φ2γ −2 = E at ( zt − µ ) = σ 2 ⇒ γ 0 = φ1γ 1 − φ2γ 2 + σ 2
1 ةB & f
وذ
k = 1: γ 1 − φ1γ 0 − φ2γ 1 = 0 ⇒ γ 1 = φ1γ 0 − φ2γ 1
k = 2 : γ 2 − φ1γ 1 − φ2γ 0 = 0 ⇒ γ 2 = φ1γ 1 − φ2γ 0
مH3K.و
k ≥ 1: γ k = φ1γ k −1 + φ2γ k −2
γ 0 7 N6
ا4! J.
ρ k = φ1 ρ k −1 + φ2 ρ k −2 , k = 1, 2,...
ρ k − φ1 ρ k −1 − φ2 ρ k −2 = 0, k = 1,2,... H3K
ا7 4J.
ا4
ا
!"دVP1. :4I5 & )
H5 ;[ام =قF. Y& H3K. O5 3!2 7;
وا4
]
ا4L & ا
ر4BوN 4
&"دO
ا
(7
%
ر اJ!
ق ا6
ا رجA ه3
و4Bوb
ت اXا
!"د
: ;
!; اوB 7
;ج ا%C 7;
وا42ار3;
ا4J26
. 4J.
ا4B "
اH%
ف1
1 − ρ0 = 1
φ1
2 − ρ1 = φ1 ρ0 + φ2 ρ −1 ⇒ ρ1 =
1 − φ2
O'&و
40
φ12
ρ 2 = φ1 ρ1 + φ2 ρ0 ⇒ ρ 2 = + φ2
1 − φ2
…r
ا اA3وه
AR ( 2 ) 4!"
7CاA
اw.
وال ا
;ا7 ه4
;
ل ا3ا
φ1 = 0.4, φ2 = 0.4 (1) H3K
ا-1
φ1 = 1.5, φ2 = −0.8 (2) H3K
ا-2
φ1 = 0.5, φ2 = −0.6 (3) H3K
ا-3
(1) H3
ACF
0.7
0.6
0.5
0.4
ACF
0.3
0.2
0.1
0.0
0 10 20
Lag
(2) H3
ACF
1.0
0.5
ACF
0.0
-0.5
0 10 20
Lag
41
(3) H3
ACF
0.5
ACF
0.0
-0.5
0 10 20
Lag
:7
;
آAR ( 2 ) 4!"
78
ا7CاA
اw. ا
;ا4
; داK
نsا
φ00 = 1, by definition
φ11 = ρ1 , by definition
1 ρ1
ρ1 ρ 2 ρ 2 − ρ12
φ22 = = ≠0
1 ρ1 1 − ρ12
ρ1 1
1 ρ1 ρ1 1 ρ1 ρ1 = φ1 + φ2 ρ1
ρ1 1 ρ2 ρ1 1 ρ 2 = φ1 ρ1 + φ2
ρ2 ρ1 ρ3 ρ 2 ρ1 ρ3 = φ1 ρ 2 + φ2 ρ1
φ33 = = =0
1 ρ1 ρ2 >0
ρ1 1 ρ1
ρ2 ρ1 1
f
A آ،7
]
اول وا2د1!"
& ا76 WآC 1 هw (
دة ا%& 7N د ا1!"
ن اf
وذ
42
1 ρ1 ⋯ ρ1 1 ρ1 ⋯ ρ1 = φ1 ρ 0 + φ2 ρ1
ρ1 1 ⋯ ρ2 ρ1 1 ⋯ ρ 2 = φ1 ρ1 + φ2 ρ 0
⋮ ⋮ ⋯ ⋮ ⋮ ⋮ ⋯ ⋮
ρ k −1 ρ k −2 ⋯ ρk ρ ρ k −2 ⋯ ρ k = φ1 ρ k −1 + φ2 ρ k −2
φkk = = k −1 = 0, k = 3, 4,...
1 ρ1 ⋯ ρ k −1 >0
ρ1 1 ⋯ ρ k −2
⋮ ⋮ ⋯ ⋮
ρ k −1 ρ k −2 ⋯ 1
إذا..
اW(
~ اb'
`2 اf
وذ
1, k =0
ρ, k =1
1
φkk = ρ 2 − ρ12
1− ρ2 , k =2
1
0, k ≥3
AR ( 2 ) 4!"
78
ا7CاA
اw.
وال ا
;ا7 ه4
;
ل ا3ا
φ1 = 0.4, φ2 = 0.4 (4) H3K
ا-4
φ1 = 1.5, φ2 = −0.8 (5) H3K
ا-5
φ1 = 0.5, φ2 = −0.6 (6) H3K
ا-6
PACF
(4) H3
0.7
0.6
0.5
PACF
0.4
0.3
0.2
0.1
0.0
0 10 20
Lag
43
(5) H3
PACF
PACF
0
-1
0 10 20
Lag
(6) H3
PACF
0.3
0.2
0.1
0.0
PACF
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
0 10 20
Lag
: ARMA(0,1) = MA(1) وq اn ا ر#
ك/\ ا0 ذج ا7 :$را
:H3K
ا7 W;3Cو
φ 0 ( B ) z t = δ + θ 1 ( B ) at
zt = δ + (1 − θ1 B ) at
zt = δ + at − θ1at −1 , at ∼ WN ( 0, σ 2 )
:78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
وداw1;!
اL1
نsا
E ( zt ) = E (δ + at − θ1at −1 ) = δ
∴µ = δ
44
ذج1!'
اW;3
و
zt − µ = at − θ1at −1
VB1;
اA وأzt −k − µ 7N 4
@ ا
!"دA`ب ه.
E ( zt − µ )( zt −k − µ ) = E ( zt −k − µ ) at − θ1 E ( zt −k − µ ) at −1 , k = 0, ±1, ±2,...
او
γ k = E ( zt −k − µ ) at − θ1 E ( zt −k − µ ) at −1 , k = 0, ±1, ±2,...
2ار3C O%.و
k = 0 : γ 0 = E ( zt − µ ) at − θ1 E ( zt − µ ) at −1
:7Cs آE ( zt − µ ) at −1 وE ( zt − µ ) at & H آL1
E ( zt − µ ) at = E ( at at ) − θ1 E ( at −1at ) = σ 2
E ( zt − µ ) at −1 = E ( at at −1 ) − θ1 E ( at −1at −1 ) = −θ1σ 2
∴γ 0 = σ 2 − θ1 ( −θ1σ 2 ) = σ 2 (1 + θ12 )
k = 1: γ 1 = E ( zt −1 − µ ) at − θ1 E ( zt −1 − µ ) at −1
γ1 −θ1
∴ γ 1 = −θ1σ 2 ⇒ ρ1 = =
γ 0 1 + θ12
1 ةJ
;[ام اF. f
وذ
k = 2 : γ 2 = E ( zt −2 − µ ) at − θ1 E ( zt −2 − µ ) at −1
∴ γ 2 = 0 ⇒ ρ2 = 0
نFN مH3K. و1 ةB & `2أ
k ≥ 2 : γ k = 0 ⇒ ρk = 0
7 7 هMA (1) 7
او4Lك & ا
ر%;!
اw1;!
ذج ا1!'
7CاA
اw. ا
;ا4
ن داFN اA3وه
:H3K
ا
45
1, k =0
−θ
ρk = 1 2 , k = 1
1 + θ1
0 k≥2
:7
;
اH3K
اO
و
θ1 = 0.8 &' -1
ACF
0.0
-0.1
-0.2
ACF
-0.3
-0.4
-0.5
0 10 20
Lag
θ1 = −0.8 &' -2
ACF
0.5
0.4
0.3
ACF
0.2
0.1
0.0
0 10 20
Lag
46
MA (1) 7
او4Lك & ا
ر%;!
اw1;!
ذج ا1!'
78
اw. ا
;ا4
; داK
نsا
φ00 = 1, by definition
φ11 = ρ1 , by definition
1 ρ1 1 ρ1
ρ1 ρ2 ρ1 0 − ρ12 −θ12 −θ12 (1 − θ12 )
φ22 = = = = =
1 ρ1 1 − ρ12 1 − ρ12 1 + θ12 + θ14 1 − θ16
ρ1 1
1 ρ1 ρ1 1 ρ1 ρ1
ρ1 1 ρ2 ρ1 1 0
ρ2 ρ1 ρ3 0 ρ1 0 ρ13 −θ13 (1 − θ12 )
φ33 = = = =
1 ρ1 ρ2 1 ρ1 0 1 − 2 ρ12 1 − θ18
ρ1 1 ρ1 ρ1 1 ρ1
ρ2 ρ1 1 0 ρ1 1
مH3K.و
−θ1k (1 − θ12 )
φkk = , k >0
1 − θ1 (
2 k +1)
:7
;
اH3K
اO
و
θ1 = −0.8 &' &' -1
PACF
0.5
0.4
0.3
0.2
PACF
0.1
0.0
-0.1
-0.2
-0.3
0 10 20
Lag
47
θ1 = 0.8 &' -2
PACF
0.0
-0.1
-0.2
PACF
-0.3
-0.4
-0.5
0 10 20
Lag
: ARMA(0,2) = MA(2) 7- اn ا ر#
ك/\ ا0 ذج ا7 :
?
:H3K
ا7 W;3Cو
φ 0 ( B ) zt = δ + θ 2 ( B ) a t
zt = δ + (1 − θ1 B − θ 2 B 2 ) at
zt = δ + at − θ1at −1 − θ 2 at −2 , at ∼ WN ( 0,σ 2 )
:78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
وداw1;!
اL1
نsا
E ( zt ) = E (δ + at − θ1at −1 − θ 2 at −2 ) = δ
∴µ = δ
ذج1!'
اW;3
و
zt − µ = at − θ1at −1 − θ 2 at −2
VB1;
اA وأzt −k − µ 7N 4
@ ا
!"دA`ب ه.
E ( zt − µ )( zt −k − µ ) = E ( zt −k − µ ) at − θ1 E ( zt −k − µ ) at −1
− θ 2 E ( zt −k − µ ) at −2 , k = 0, ±1, ±2,...
او
48
γ k = E ( zt −k − µ ) at − θ1E ( zt −k − µ ) at −1 − θ 2 E ( zt −k − µ ) at −2 , k = 0, ±1, ±2,...
2ار3C O%.و
γ 0 = (1 + θ12 + θ 22 ) σ 2
γ 1 = ( −θ1 + θ1θ 2 ) σ 2
γ 2 = −θ 2σ 2
γ k = 0, k > 2
γ 0 7 4! J
.و
−θ1 + θ1θ 2
ρ1 =
1 + θ12 + θ 22
−θ 2
ρ2 =
1 + θ12 + θ 22
ρ k = 0, k > 2
4
داH3 7 W;3Cو
1, k =0
−θ + θ θ
1 2 1 22 , k = 1
1 + θ1 + θ 2
ρk =
−θ 2 , k =2
1 + θ12 + θ 22
0, k >2
MA ( 2 ) 4!"
7CاA
اw.
وال ا
;ا7 ه4
;
ل ا3ا
θ1 = 0.4, θ 2 = 0.4 (7) H3K
ا-7
θ1 = 1.5, θ 2 = −0.8 (8) H3K
ا-8
θ1 = 0.5, θ 2 = −0.6 (9) H3K
ا-9
49
(7) H3
ACF
0.0
-0.1
ACF -0.2
-0.3
0 10 20
Lag
(8) H3
ACF
0.2
0.1
0.0
-0.1
ACF
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
0 10 20
Lag
(9) H3
ACF
0.4
0.3
0.2
0.1
ACF
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
0 10 20
Lag
50
& ك%;!
اw1;!
ذج ا1!'
78
ا7CاA
اw. ا
;ا4
اY& H3 د2ا إL W"
& ا
)J
2ار3C O! ورO. %
ب11 e2"C ف
;[م1 اAO
وMA ( 2 ) 4
]
ا4Lا
ر
:4
;
ا
!"
) ا
θ1 = 0.4, θ 2 = 0.4 (10) H3K
ا -10
θ1 = 1.5, θ 2 = −0.8 (11) H3K
ا -11
θ1 = 0.5, θ 2 = −0.6 (12) H3K
ا -12
(10) H3
PACF
0.0
-0.1
PACF
-0.2
-0.3
0 10 20
Lag
(11) H3
PACF
0.2
0.1
0.0
-0.1
PACF
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
0 10 20
Lag
51
(12) H3
PACF
0.4
0.3
0.2
0.1
PACF
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
0 10 20
Lag
: ARMA(1,1) n ا ر#
ا9 ار ا/7jا-
ك/\ ا0 ذج ا7 :0د0
:H3K
اW;32و
φ1 ( B ) zt = δ + θ1 ( B ) at
(1 − φ1B ) zt = δ + (1 − θ1B ) at
zt − φ1 zt −1 = δ + at − θ1at −1
zt = δ + φ1 zt −1 + at − θ1at −1 , at ∼ WN ( 0, σ 2 ) , φ1 ≠ θ1
&; خ9 ! ط ا2 , وه'ك طθ1 < 1 بJ 9 وط اφ1 < 1 ارJ;9ط ا
ذج1!
ذج إ1!'
`! م إ&; خ ا2 طK
ا اA وهφ1 ≠ θ1 1 وهDegeneracy Condition
4! J
.( و1 − φ1B ) zt = δ + (1 − θ1B ) at 4B "
! اN φ1 = θ1 ن1 آ4
5 7bN 4L درHBأ
δ
ARMA ( 0, 0 ) 1 وهδ ′ = Q5 zt = δ ′ + at ^( 2 ذج1!'
(
أن ا1 − φ1B )
1 − φ1
:7
;
آw1;!
اL1
(1 − φ1B ) zt = δ + (1 − θ1B ) at
zt =
δ
+
(1 − θ1B ) a
1 − φ1 (1 − φ1 B )
t
E ( zt ) =
δ (1 − θ1B ) E a
+ ( t)
1 − φ1 (1 − φ1 B )
اA3 وهφ1 < 1 نf
وذ
δ
E ( zt ) =
1 − φ1
52
δ
δ \21";
. وδ = µ (1 − φ1 ) أوE ( zt ) = µ = أي
1 − φ1
zt = µ (1 − φ1 ) + φ1 zt −1 + at − θ1at −1
( zt − µ ) − φ1 ( zt −1 − µ ) = at − θ1at −1
N6
VB1;
اA ( وأzt −k − µ ) , k = 0, ±1, ±2,... %
. 4
ا
!"د7N= `ب.و
E ( zt −k − µ )( zt − µ ) − φ1 E ( zt −k − µ )( zt −1 − µ ) = E ( zt −k − µ ) at − θ1 E ( zt −k − µ ) at −1 ,
k = 0, ±1, ±2,...
O'&و
γ k − φ1γ k −1 = E ( zt −k − µ ) at − θ1E ( zt −k − µ ) at −1 , k = 0, ±1, ±2,...
2ار3C O%.و
k = 0 γ 0 − φ1γ 1 = E ( zt − µ ) at − θ1 E ( zt − µ ) at −1
4B "
`ب ا. E ( zt − µ ) at −1 وE ( zt − µ ) at & Hن آs اL1
( zt − µ ) − φ1 ( zt −1 − µ ) = at − θ1at −1
VB1;
اA وأat −1 وat & H آ7N
E ( zt − µ ) at − φ1 E ( zt −1 − µ ) at = E [ at at ] − θ1 E [ at −1at ]
1 ةJ
و& ا
E ( zt − µ ) at − φ1 ( 0 ) = σ 2 − θ1 ( 0 )
E ( zt − µ ) at = σ 2
و
E ( zt − µ ) at −1 − φ1E ( zt −1 − µ ) at −1 = E [at at −1 ] − θ1E [ at −1at −1 ]
E ( zt − µ ) at −1 − φ1σ 2 = 0 − θ1σ 2
∴ E ( zt − µ ) at −1 = σ 2 (φ1 − θ1 )
4J.
ا4Y
ا7N \21";
.و
k = 0 γ 0 − φ1γ 1 = σ 2 − θ1σ 2 (φ1 − θ1 )
∴γ 0 − φ1γ 1 = σ 2 1 − θ1 (φ1 − θ1 )
و
53
k = 1 γ 1 − φ1γ 0 = E ( zt −1 − µ ) at − θ1 E ( zt −1 − µ ) at −1
∴γ 1 − φ1γ 0 = −θ1σ 2
و
k = 2 γ 2 − φ1γ 1 = E ( zt −2 − µ ) at − θ1 E ( zt − 2 − µ ) at −1 = 0
∴ k ≥ 2 γ k − φ1γ k −1 = 0
تXو& ا
!"د
γ 0 − φ1γ 1 = σ 2 1 − θ1 (φ1 − θ1 )
و
γ 1 − φ1γ 0 = −θ1σ 2
1 + θ12 − 2φ1θ1 2
γ0 = σ
1 − φ12
γ1 =
(1 − φ1θ1 )(φ1 − θ1 ) σ 2
1 − φ12
;J.
; اB "
و& ا
γ 1 (1 − φ1θ1 )(φ1 − θ1 )
ρ1 = =
γ0 1 + θ12 − 2φ1θ1
4B "
و& ا
γ k − φ1γ k −1 = 0, k ≥ 2
γ 0 4! J
.و
ρ k − φ1 ρ k −1 = 0, k ≥ 2
وρ0 = 1 4
) اوJ
;[ام اF. k ≥ 2 )B V!
2ار3C 4
@ ا
!"دA هH5 3!2و
]!N ρ1 =
(1 − φ1θ1 )(φ1 − θ1 )
1 + θ12 − 2φ1θ1
ρ 2 = φ1 ρ1
ρ 2 = φ1
(1 − φ1θ1 )(φ1 − θ1 )
1 + θ12 − 2φ1θ1
ρ 3 = φ1 ρ 2
54
ρ 3 = φ12
(1 − φ1θ1 )(φ1 − θ1 )
1 + θ12 − 2φ1θ1
.اA3وه
H3K
اARMA (1,1) ذج1!'
7CاA
اw. ا
;ا4
داW;3
1, k =0
(1 − φ1θ1 )(φ1 − θ1 )
ρk = , k =1
1 + θ 1
2
− 2φ 1θ 1
φ1 ρ k −1 k≥2
φ1 = 0.9,θ1 = −0.5 )J
7CاA
اw. ا
;ا4
دا76"2 13 H3
(13)H3
A C F o f A R M A (1 ,1 )
1 .0
0 .9
0 .8
0 .7
0 .6
C1
0 .5
0 .4
0 .3
0 .2
0 .1
0 .0
0 5 10 15
Lag
φ1 = −0.9,θ1 = −0.5 )J
7CاA
اw. ا
;ا4
دا76"2 14 H3
(14)H3
A C F o f A R M A ( 1 ,1 )
0 .5
C1
0 .0
-0 .5
0 5 10 15
Lag
55
. أو &;دد5 @ واC إ7N ;[& اC ARMA (1,1) ذج1!'
7CاA
اw. ا
;ا4
ان دا5
&ى انAR (1) ذج1!'
7CاA
اw. ا
;ا4
!& داC *(KC اA ه7N 7 وه4(
وا4(L1!
) اJ
ا
( ρ k = φ1k −1 ρ1 , k ≥ 2 ه أن. ) ρ1 & (أ2 &[;
ا
:7
;
ب آ11 أو11 e2"C & W %C φkk 78
ا7CاA
اw. ا
;ا4
دا
2ار3C φkk L1 ب11 e2"C &
φ00 = 1, by definition
φ11 = ρ1 =
(1 − φ1θ1 )(φ1 − θ1 )
1 + θ12 − 2φ1θ1
ρ 2 − φ11 ρ1
φ22 =
1 − φ11 ρ1
ρ −φ ρ −φ ρ
φ33 = 3 21 2 22 1 , φ21 = φ11 − φ22φ11
1 − φ21 ρ1 − φ22 ρ 2
.2ار3C )J
ا4J. W %C اA3وه
φ1 = 0.9,θ1 = −0.5 )J
]!N
φ11 = 0.944186 φ22 = -0.384471 φ33 = 0.183710
φ44 = -0.908462 φ55 = 0.452979 φ66 = -0.226337
φ77 = 0.113154 φ88 = -0.565702 φ99 = 0.282834
15 H3 7N )J
@ اAو
) ه
15 H3
P A C F o f A R M A (1 ,1 )
1 .0
0 .5
C2
0 .0
0 5 10 15
Lag
φ1 = −0.9,θ1 = −0.5 )J
78
ا7CاA
اw. ا
;ا4
( دا2 16 H3
56
16 H3
P A C F o f A R M A (1 ,1 )
0 .3
0 .2
0 .1
0 .0
-0 .1
C2
-0 .2
-0 .3
-0 .4
-0 .5
-0 .6
0 5 10 15
Lag
أو &;دد5 @ واC إ7N ;[& اC ARMA (1,1) ذج1!'
78
ا7CاA
اw. ا
;ا4
ان دا5
MA (1) ذج1!'
78
ا7CاA
اw. ا
;ا4
!& داC *(KC اA ه7N 7 وه4(
وا4(L1!
) اJ
ا.
. φ11 = ρ1 4
او4!J
" ا. (أ2 &[;
&ى ان ا
: ARMA(p,q) ذج7 > ? اص:
AR(p) ذج7 :rأو
:7
;
. !;2و
.4(
وا
;[&ات ا4X & ا
;[&ات اw & ن13;C و8O X ;!C 7C ذاw.اC 4
دا-1
أيk > p تb[;
) اJ
رb+ن & أ13;C 78L 7C ذاw.اC 4
دا-2
φ11 = φ22 = φ33 = ⋯ = φ pp ≠ 0
φ p +1, p +1 = φ p + 2, p + 2 = ⋯ = 0
. k > p e[;
" ا. 78
ا7CاA
اw. ا
;ا4
دا7N "6B اA ! ه2و
: MA(q) ذج7 :7]
:7
;
. !;2و
أيk > q تb[;
) اJ
رb+ن & أ13;C 7C ذاw.اC 4
دا-1
ρ1 = ρ 2 = ρ 3 = ⋯ = ρ q ≠ 0
ρ q+1,q+1 = ρ q+ 2,q +2 = ⋯ = 0
. k > q e[;
" ا. 7CاA
اw. ا
;ا4
دا7N "6B اA ! ه2و
57
-2دا
C 4ا w.ذا 8O
X ;!C 78L 7Cو13;Cن & & wا
;[&ات ا 4Xوا
;[&ات
ا
(.4
5Xا9زدوا1!
. Duality 4Lذ AR 7Lو .MA
] :-ا ذج ا:ARMA(p,q) \:
و:7
;
. !;2
دوال ا
;ا w.ا
Aا 7Cوا
;ا w.ا
Aا 7Cا
1!'
78ذج ا
![; 8O X ;!C wو13;Cن & & w
ا
;[&ات ا 4Xوا
;[&ات ا
( 4ا
; 7O;'C 7إ
ا
bآ! زاد ا
;[13C &' . k eن
FN k > q − pن دا
4ا
;ا w.ا
Aا%;C 7Cد & Lء ا% 9ار ا
Aا1!'
7Cذج و '& 13Cن
FN k > p − qن دا
4ا
;ا w.ا
Aا 7Cا
%;C 78د & Lء ا
!; w1ا
!;%ك
'!1ذج.
58
ﺍﻟﻔﺼﻞ ﺍﻟﺜﺎﻟﺚ
:Nonstationar Time Series Models
ا
ةW
ت ا,ذج ا7
:\0 اa
ار0j م ا:rاو
ارJ;
اول طK
ا ان
ى 4'&ز 4 ;& ارJ;9 6 e2"C &
]!N ، &
ل ا1= S.U 4 ;!
اw1;& ن132 أنW6;2 E ( zt ) = µ = constant ∀t
76[
اف ا9ذج ا1!'
zt = b0 + b1t + at , at ∼ WN ( 0, σ 2 ) , b0 , b1 ∈ ( −∞, ∞ )
1 هw1;!
ان ا
E ( z )t = b0 + b1t
.4
%
@ اA ه7N J%;& z ار اولJ;9 اي ان ط ا، &
4( '
. S.U z 1وه
ذج1!'
ا2b;
اH& (6;. f
∇ وذzt H21%;
ول ا%'
wt = ∇zt = zt − zt −1 = b0 + b1t + at − b0 − b1 ( t − 1) − at −1
=b1 + at − at −1 = b1 + ct
∴ wt = b1 + ct , ct ∼ WN ( 0,ν 2 )
( σ 2 وν 2 . 4B "
اL أو: 2!C )
wt ة2
ا4 ;!
اw1;&
نsا
E ( wt ) = b1 = constant ∀t
اول2b;
اAة ) أي أJ; !
اz 4 ;!
( = ∇ ا1 − B ) H21%;
( ا6C أي ان
.ةJ; & 4 ;&
إO
15 (4 ;!
7".;
اف ا9ذج ا1!
, آ!]ل
zt = b0 + b1t + b2t 2 + at , at ∼ WN ( 0, σ 2 ) , b0 , b1 , b2 ∈ ( −∞, ∞ )
w1;!
د ا2F.
E ( zt ) = b0 + b1t + b2t 2
(7
]
ا2b;
اA∇ ) أ2 zt H21%;
اA{. .J; & z ذج1!'
أي ان ا، &
";! ا2 1وه
59
∇ 2 zt = ∇2 ( b0 + b1t + b2t 2 + at )
(1 − 2 B + B ) z = (1 − 2 B + B )( b
2
t
2
0 + b1t + b2t 2 + at )
wt = {b0 − 2b0 + b0 } + {b1t − 2b1 ( t − 1) + b1 ( t − 2 )} +
{b t
2
2
− 2b2 ( t − 1) + b2 ( t − 2 ) +
2 2
}
{at − 2at −1 + at −2 }
= 2b2 + {at − 2at −1 + at −2 }
=b′ + ht , ht ∼ WN ( 0,τ 2 )
اA3وه
wt = ∇ 2 zt = b′ + ht , ht ∼ WN ( 0,τ 2 )
E ( wt ) = b′ = constant ∀t
اO
15 ةJ; !
اz 4 ;!
( ا7
]
ا2b;
اA∇ )أي ا2 H21%;
( ا6C أي ان
.ةJ; &
.( σ 2 وτ 2 . 4B "
اL أو: 2!C )
H3K
اJ; !
اz ذج1!'
م إذا آن اH3K.
zt = b0 + b1t + ⋯ + bd t d + at , at ∼ WN ( 0,σ 2 ) , b0 , b1 ,⋯ , bd ∈ ( −∞, ∞ )
.J; & ذج1! 1 هwt = ∇d zt أي ان،J; & ذج1!
* إ1%2 ∇d zt H21%;
ن اFN
:16
\0 اa
ا
ةW
zt = b0 + b1t + ⋯ + bd t d + at , at ∼ WN ( 0,σ 2 ) , b0 , b1 ,⋯ , bd ∈ ( −∞, ∞ )
. إ'
ةL / d n
= رY∇ وه اd zt . /ا
:# اa ار0j م ا:ً 7]
7 ]
ط اK
ا،4'& ز4 ;& ارJ;9 6 e2"C &
V ( zt ) = γ 0 = constant ∀t
. t )B V!
S.U 2(;
ن ا132 أنW6;2
78ا1K"
ا7K!
ذج ا1!'
]!N
60
zt = zt −1 + at , at ∼ WN ( 0,σ 2 )
ر3;!
\ ا21";
& ا
zt = a1 + a2 + ⋯ + at
2(;
واVB1;
اAF.و
E ( zt ) = 0 = constant ∀t
V ( zt ) = tσ 2
. t &
";! ا2 2(;
أن ا5
و
اول2b;
اA{.
wt = ∇zt = zt − zt −1 = at
2(;
واVB1;
اAF.و
E ( wt ) = 0 = constant ∀t
V ( wt ) = σ 2 = constant ∀t
.ةJ; & 4 ;&
إ2(;
ا7N ةJ; !
اz 4 ;!
ل ا15 اول2b;
إذًا ا
H3K
اY;& (w1;&) ى1; !
4
دا2(;
آن ا1
مH3K.
V ( zt ) = cf ( µt )
Y;2 w1;& ى أو1; & µt و4(
z 4!B 6"C 4N &"و4
داf (⋅) وS.U c > 0 Q5
4
د دا2 أي إT ( zt ) H21%C د2ول إ%
'";! ا
& وه2 2(;
ن اFN 7
;
. ا
& وV&
. 2(;
ار اJ;9 T (⋅)
H21%;
ا
ztλ − 1
y t = T ( zt ) =
λ
76"2 7
;
ا
ول ا.H21%;
&") ا1 هλ ∈ ( −∞, ∞ ) Q5 2(;
ا7N ةJ; & 4 ;& 76"2
:O
4.J!
ت ا21%;
اV& λ )"!
&) اآ] إ;[اJ
ا
λ -0.1 -0.5 0.0 0.5 1.0
yt 1 1 ln zt zt zt
zt zt
61
:ل-
zt 2(;
واw1;!
ا7N ةJ; & z 4 ;!
()اH3K
ا
O r ig in a l S e r ie s
400
300
z(t)
200
100
In d e x 10 20 30 40 50 60 70 80 90
yt = ln zt H21%;
اء اLF. 2(;
اS(]C ". 4 ;!
)ب( اH3K
ا
T r a n s f o r m e d S e r ie s
6 .0
5 .5
ln z(t)
5 .0
In d e x 10 20 30 40 50 60 70 80 90
∇yt = yt − yt −1 اول2b;
اء اL" إ. yt 4
1%!
ا4 ;!
)ج( اH3K
ا
D if f e re n c e d a n d T ra n s f o rm e d S e rie s
0 .2
0 .1
y(t)-y(t-1)
0 .0
-0 .1
-0 .2
In d e x 10 20 30 40 50 60 70 80 90
. 2(;
واw1;!
& اH آ7N ةJ; & 4 ;!
اS%(+ اe آ5X
62
Autoregressive- (p,d,q) n ا ر#
ك/\ ا0 ا-
)ا-ا9 ار ا/7jذج ا7
Integrated-Moving Average Models ARIMA(p,d,q)
& ك%;& w1;&-7Cار ذا%
ذج أ1!
H3 wt = ∇d zt ةJ; !
ا4 ;!
ا4LA!
3!2
:7
;
( آp, q ) 4Lا
ر
φ p ( B ) wt = φ p ( B ) ∇d zt = δ + θ q ( B ) at , at ∼ WN ( 0, σ 2 )
أو
φ p ( B )(1 − B ) zt = δ + θ q ( B ) at , at ∼ WN ( 0, σ 2 )
d
( p, d , q ) 4Lك & ا
ر%;!
اw1;!
ا-7&3;
ا-7CاA
ار ا%
9ذج ا1!
! 2 ذج1!'
ا اAوه
.
اف9 &") اδ ∈ ( −∞, ∞ ) Q5
: ( p, d , q ) 4Lك & ا
ر%;!
اw1;!
ا-7&3;
ا-7CاA
ار ا%
9
!ذج ا4]&أ
: ARIMA(1,1,0) =ARI(1,1) ( أو1,1) n ا ر#
)ا-ا9 ارا/7j ذج ا7 :rاو
H3K
اW;32و
φ1 ( B )(1 − B ) zt = δ + θ 0 ( B ) at , at ∼ WN ( 0,σ 2 )
(1 − φ1B )(1 − B ) zt = δ + at
{1 − (φ1 + 1) B + φ1B 2 } zt = δ + at
أي
zt = δ + (φ1 + 1) zt −1 − φ1 zt −2 + at , at ∼ WN ( 0, σ 2 ) , φ1 < 1
(1,1) n ا ر#
ك/\ ا0 ا-
) ذج ا7 :7]
: ARIMA(0,1,1) = IMA(1,1) أو
H3K
اW;32و
φ0 ( B )(1 − B ) zt = δ + θ1 ( B ) at , at ∼ WN ( 0, σ 2 )
(1 − B ) zt = δ + (1 − θ1B ) at , at ∼ WN ( 0, σ 2 ) , θ1 < 1
zt − zt −1 = δ + at − θ1at , at ∼ WN ( 0, σ 2 ) , θ1 < 1
أي
zt = δ + zt −1 + at − θ1at , at ∼ WN ( 0,σ 2 ) , θ1 < 1
63
أوRandom Walk with Trend Model
افV7c$ ذج ا ا ا7 :-]
: ARIMA(0,1,0)
H3K
اW;32و
φ0 ( B )(1 − B ) zt = δ + θ 0 ( B ) at , at ∼ WN ( 0, σ 2 )
(1 − B ) zt = δ + at , at ∼ WN ( 0,σ 2 )
أي
zt = δ + zt −1 + at , at ∼ WN ( 0,σ 2 )
:ARMA(p,q) ذج7 .- وψ ( B ) وزانqدا ا
H3K
اARMA ( p, q ) ( أن آ;('
!ذج
φ p ( B ) zt = δ + θ q ( B ) at , at ∼ WN ( 0, σ 2 )
w1;!
اف ا%
9 اH3K. أو
φ p ( B )( zt − µ ) = θ q ( B ) at , at ∼ WN ( 0,σ 2 )
φ p ( B ) 7CاA
ار ا%
9 اH& 4! J
. ;
%
آ; ا7N
δ θq ( B )
zt = + at , at ∼ WN ( 0, σ 2 )
φ p (1) φp (B)
θq ( B )
zt − µ = at , at ∼ WN ( 0, σ 2 )
φp (B)
θq ( B )
φ p ( B ) = 0 ورAL نf
وذ4.رJ;& 4 & H3KC 4( '
ة اJ; !
ان
'!ذج ا5X
φp (B)
δ
7N w1;!
اف ا%
9 اH3K. 7b;3
ف1 اAO
و = µ `2ة ا51
ة ا8 رج داVJC
φ p (1)
4
;
;' اKB'&
θq ( B )
zt − µ = at , at ∼ WN ( 0,σ 2 )
φp ( B)
4.رJ;!
ا4 !
ا
64
θq ( B )
ψ ( B) =
φp (B)
H3K
اW;3C 7;
وا
θq ( B)
ψ ( B) = = ψ 0 B 0 + ψ 1 B1 + ψ 2 B 2 + ψ 3 B 3 + ⋯ , ψ 0 = 1
φp (B)
. اوزان4
! داC
:17
I,$ ';
اARMA ( p, q ) وزان ذجqدا ا
θq ( B)
ψ ( B) = = ψ 0 B 0 + ψ 1 B1 + ψ 2 B 2 + ψ 3 B 3 + ⋯ , ψ 0 = 1
φp (B)
θq ( B) ∞
ψ ( B) = = ∑ψ j B j , ψ 0 = 1
φ p ( B ) j =0
ψ 0 = 1,ψ 1 ,ψ 2 ,ψ 3 ,⋯ وزان هq اJK
θq ( B )
H]!C ! 2 zt − µ = at , at ∼ WN ( 0,σ 2 ) H3K
ي اA
ذج ا1!'
ا: !K,
φp ( B)
. ARMA ( p, q )
'!ذج78O
ك ا%;!
اw1;!
ا
: اذجv وزانq ا ا-
أ
: AR(1) وزان ذجqدا ا
H3K
اW;32 AR(1) ذج1!
φ1 ( B )( zt − µ ) = θ 0 ( B ) at , at ∼ WN ( 0,σ 2 )
(1 − φ1B )( zt − µ ) = at
1
zt − µ = a
(1 − φ1B ) t
z t − µ = ψ ( B ) at
Q5
1
ψ (B) =
(1 − φ1B )
65
4
;
ا4J26
. ψ 1 ,ψ 2 ,ψ 3 ,⋯ اوزانL1
ف1
1
ψ ( B) =
(1 − φ1B )
ψ ( B )(1 − φ1B ) ≡ 1
(1 + ψ B + ψ
1 2 B 2 + ψ 3 B 3 + ⋯) (1 − φ1 B ) ≡ 1
.42 &; و4B "
ا7N= B j أي ان &"& ت:N3C 4B 7 اة ه4B "
أن ا5X
4B "
ا7N= B j ! واة &"& ت.و
(1 + ψ B + ψ
1 2 B 2 + ψ 3 B 3 + ⋯) (1 − φ1 B ) ≡ 1, φ1 < 1
B 0 : (1)(1) ≡ 1
B1 : ψ 1 − φ1 ≡ 0 ⇒ ψ 1 = φ1
B 2 : ψ 2 − ψ 1φ1 ≡ 0 ⇒ ψ 2 = ψ 1φ1 = φ12
B 3 : ψ 3 − ψ 2φ1 ≡ 0 ⇒ ψ 3 = ψ 2φ1 = φ13
⋮
B j : ψ j − ψ j −1φ1 ≡ 0 ⇒ ψ j = ψ j −1φ1 = φ1j
7 هAR(1) ذج1!'
أي ان اوزان
ψ j = φ1j , φ1 < 1
: MA(1) وزان ذجqدا ا
H3K
اW;32 MA(1) ذج1!
φ0 ( B )( zt − µ ) = θ1 ( B ) at , at ∼ WN ( 0,σ 2 )
( zt − µ ) = (1 − θ1B ) at
z t − µ = ψ ( B ) at
Q5
ψ ( B ) = (1 − θ1B )
4B "
ا7N= B j ! واة &"& ت.
ψ 1 = −θ1 , ψ 2 = ψ 3 = ⋯ = 0
أي
1, j=0
ψ j = −θ1 , j =1
0, j≥2
66
: AR(2) وزان ذجqدا ا
H3K
اW;32 AR(2) ذج1!
φ2 ( B )( zt − µ ) = θ 0 ( B ) at , at ∼ WN ( 0, σ 2 )
(1 − φ B − φ B ) ( z − µ ) = a
1 2
2
t t
1
zt − µ = a
(1 − φ1B − φ2 B 2 ) t
z t − µ = ψ ( B ) at
Q5
1
ψ ( B) =
(1 − φ1B − φ2 B 2 )
4J.
ا4J26
. ψ 1 ,ψ 2 ,ψ 3 ,⋯ اوزانL1
و
ψ ( B ) (1 − φ1B − φ2 B 2 ) ≡ 1
(1 + ψ B + ψ
1 2 B 2 + ψ 3 B 3 + ⋯)(1 − φ1 B − φ2 B 2 ) ≡ 1
B1 : ψ 1 − φ1 = 0 ⇒ ψ 1 = φ1
B 2 : ψ 2 − φψ
1 1 − φ 2 = 0 ⇒ ψ 2 = φψ
1 1 + φ 2 = φ1 + φ2
2
B 3 : ψ 3 − φψ
1 2 − φ 2ψ 1 = 0 ⇒ ψ 3 = φψ
1 2 + φ2ψ 1
⋮
B j : ψ j − φψ
1 j −1 − φ2ψ j − 2 = 0 ⇒ ψ j = φψ
1 j −1 + φ 2ψ j − 2
7 هAR(2) ذج1!'
أي ان اوزان
1, j=0
φ , j =1
1
ψj = 2
φ1 + φ2 , j=2
φψ
1 j −1 + φ2ψ j − 2 , j≥3
: MA(2) وزان ذجqدا ا
H3K
اW;32 MA(2) ذج1!
φ0 ( B )( zt − µ ) = θ 2 ( B ) at , at ∼ WN ( 0, σ 2 )
( zt − µ ) = (1 − θ1B − θ 2 B 2 ) at
z t − µ = ψ ( B ) at
Q5
ψ ( B ) = (1 − θ1B − θ 2 B 2 )
67
4B "
ا7N= B j ! واة &"& ت.
ψ 1 = −θ1 , ψ 2 = −θ 2 , ψ 3 = ψ 4 = ψ 5 ⋯ = 0
أي
1, j=0
−θ , j =1
ψj = 1
−θ 2 , j=2
0, j≥2
: ARMA(1,1) وزان ذجqدا ا
H3K
اW;32 ARMA(1,1) ذج1!
φ1 ( B )( zt − µ ) = θ1 ( B ) at , at ∼ WN ( 0,σ 2 )
(1 − φ1B )( zt − µ ) = (1 − θ1B ) at
zt − µ =
(1 − θ1B ) a
(1 − φ1B ) t
z t − µ = ψ ( B ) at
Q5
ψ (B) =
(1 − θ1B )
(1 − φ1B )
4J.
ا4J26
. ψ 1 ,ψ 2 ,ψ 3 ,⋯ اوزانL1
و
ψ ( B )(1 − φ1B ) ≡ (1 − θ1B )
(1 + ψ B + ψ
1 2 B 2 + ψ 3 B 3 + ⋯) (1 − φ1B ) ≡ (1 − θ1 B )
B1 : ψ 1 − φ1 = −θ1 ⇒ ψ 1 = φ1 − θ1
B 2 : ψ 2 − φψ 1 1 = φ1 (φ1 − θ1 )
1 1 = 0 ⇒ ψ 2 = φψ
B 3 : ψ 3 − φψ 1 2 = φ1 (φ1 − θ1 )
1 2 = 0 ⇒ ψ 3 = φψ
2
⋮
B j : ψ j − φψ
1 j −1 = 0 ⇒ ψ j = φψ
1 j −1 = φ1
j −1
(φ1 − θ1 )
7 هARMA(1,1) ذج1!'
أي ان اوزان
ψ j = φψ
1 j −1 = φ1
j −1
(φ1 − θ1 ) , j ≥ 1, φ1 < 1, φ1 ≠ θ1
68
: ARI(1) وزان ذجqدا ا
H3K
اW;32 ARI(1) ذج1!
φ1 ( B )(1 − B )( zt − µ ) = at , at ∼ WN ( 0, σ 2 )
1
zt − µ = a
(1 − φ1B )(1 − B ) t
z t − µ = ψ ( B ) at
Q5
1
ψ ( B) =
(1 − φ1B )(1 − B )
4J.
ا4J26
. ψ 1 ,ψ 2 ,ψ 3 ,⋯ اوزانL1
و
ψ ( B )(1 − φ1B )(1 − B ) ≡ 1
(1 + ψ B + ψ
1 2 B 2 + ψ 3B 3 + ⋯) (1 − φ1B )(1 − B ) ≡ 1
(1 + ψ B + ψ
1 2 B 2 + ψ 3B 3 + ⋯) (1 − (φ1 + 1) B + φ1B 2 ) ≡ 1
B1 : ψ 1 − (φ1 + 1) = 0 ⇒ ψ 1 = φ1 + 1
B 2 : ψ 2 − (φ1 + 1)ψ 1 + φ1 = 0 ⇒ ψ 2 = (φ1 + 1)ψ 1 + φ1 = (φ1 + 1) + φ1
2
B 3 : ψ 3 − (φ1 + 1)ψ 2 + φψ
1 1 = 0 ⇒ ψ 3 = (φ1 + 1)ψ 2 − φψ
1 1
⋮
B j : ψ j − (φ1 + 1)ψ j −1 + φψ
1 j − 2 = 0 ⇒ ψ j = (φ1 + 1)ψ j −1 − φψ
1 j −2
7 هARI(1) ذج1!'
أي ان اوزان
1, j=0
φ + 1, j =1
1
ψj =
(φ1 + 1) + φ1 , j=2
2
(φ1 + 1)ψ j −1 − φψ
1 j −2 , j≥3
L1
وأا
ARIMA(1,0,1) أوRandom Walk Mdel وزان ذج ا ا اqدا ا
H3K
اW;32و
zt = zt −1 + at , at ∼ WN ( 0,σ 2 )
أي
69
zt − zt −1 = at , at ∼ WN ( 0,σ 2 )
(1 − B ) zt = at
1
zt = a
(1 − B ) t
Q5
1
ψ ( B) =
(1 − B )
4J.
ا4J26
. ψ 1 ,ψ 2 ,ψ 3 ,⋯ اوزانL1
و
ψ ( B )(1 − B ) ≡ 1
(1 + ψ B + ψ
1 2 B 2 + ψ 3 B 3 + ⋯) (1 − B ) ≡ 1
B1 : ψ 1 − 1 = 0 ⇒ ψ 1 = 1
B 2 :ψ 2 −ψ 1 = 0 ⇒ ψ 2 = ψ 1 = 1
B 3 :ψ 3 −ψ 2 = 0 ⇒ ψ 3 = ψ 2 = 1
⋮
B j : ψ j − ψ j −1 = 0 ⇒ ψ j = ψ j −1 = 1
7 هARIMA ( 0,1, 0 ) 78ا1K"
ا7K!
ذج ا1!'
أي ان اوزان
ψ j = 1, j ≥1
: ψ ( B ) وزانq ? اص دا اv$
H3K
اARMA(p,q) 4Lك & ا
ر%;!
اw1;!
ا-7CاA
ار ا%
9ذج ا1!
'(;( أن آ
zt − µ = ψ ( B ) at , at ∼ WN ( 0,σ 2 )
H3K
* اB "
@ اA ه4.;3.و
zt − µ = at + ψ 1at −1 + ψ 2at −2 + ψ 3at −3 + ⋯
∞
= ∑ψ j at − j , ψ 0 = 1
j =0
∞
:4
;
ا42I'
(ت اU إ3!2 *
FN ∑ j =0
ψ 2j < ∞ رب ايJ;C ' ان اوزانP;Nوإذا ا
70
:1
!7
N) ي9 ا
واARMA(p,q) n ا ر#
ك/\ ا0 ا-ا9 ار ا/7j ذج ا
.) ' ا
∞
zt − µ = ∑ψ j at − j , at ∼ WN ( 0, σ 2 ) , ψ 0 = 1, ∑ j =0ψ 2j < ∞
∞
j =0
\ ه0 ا-1
E ( zt ) = µ , ∀t
I,$ '; ا9\ ا$ دا ا ا-2
∞
∑ψ ψ j j+k
ρk = j =0
∞ , k = 0,1, 2,⋯
∑ψ 2j
j =0
اوزان4
داJ.
L وAR(1) 4L & ا
ر7CاA
ار ا%
9ذج ا1!'
:ل-
ψ j = φ1j , φ1 < 1
7CاA
اw. ا
;ا4
دا
∞ ∞
φ1k
∑ψ ψ j j +k ∑φ φ j j+k
1 1
1 − φ12
ρk = j =0
∞ = j =0
∞ = = φ1k , k = 0,1, 2,⋯
1
∑ψ
j =0
2
j ∑φ j =0
1
2j
1 − φ12
4J.
ا4 ;'
~ اb
7وه
4
;
'!ذج ا7CاA
اw. دوال ا
;اL( أو2) 2 42I
{;[ام. :#
AR(2), MA(1), MA(2), ARMA(1,1), ARMA(2,1), ARMA(1,2)
71
ﺍﻟﻔﺼﻞ ﺍﻟﺮﺍﺑﻊ
ا4ات ذات
%$
\0ا g;:اqد '7ذج )ARMA(p,q
Minimum Mean Square Error Forecasts for ARMA(p,q) Models
ﻓﻲ ﺍﻝﻔﻘﺭﺓ ﺍﻝﺴﺎﺒﻘﺔ ﻜﺘﺒﻨﺎ ﻨﻤﻭﺫﺝ ﺍﻹﻨﺤﺩﺍﺭ ﺍﻝﺫﺍﺘﻲ-ﺍﻝﻤﺘﻭﺴﻁ ﺍﻝﻤﺘﺤﺭﻙ ﻤﻥ ﺍﻝﺩﺭﺠﺔ )ARMA(p,q
ﺍﻝﻤﺴﺘﻘﺭ ﻋﻠﻰ ﺍﻝﺸﻜل
∞
∞ < zt − µ = ∑ψ j at − j , at ∼ WN ( 0, σ 2 ) , ψ 0 = 1, ∑ j =0ψ 2j
∞
j =0
ﺃﻭ
⋯ zt − µ = at + ψ 1at −1 + ψ 2at −2 + ψ 3at −3 +
∞
= ∑ψ j at − j , ψ 0 = 1
j =0
ﻤﻼﺤﻅﺔ :ﻫﺫﺍ ﻴﻨﻁﺒﻕ ﺃﻴﻀﺎ ﻋﻠﻰ ﻨﻤﺎﺫﺝ ) ARIMA(p,d,qﺒﺸﻜل ﻋﺎﻡ.
ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ﺍﻝﺘﻨﺒﺅﺍﺕ zn ( ℓ ) , ℓ ≥ 1 ﻝﻤﺘﺴﻠﺴﻠﺔ ﺯﻤﻨﻴﺔ ﻤﺸﺎﻫﺩﺓ } {z1 , z2 ,⋯ , zn−1 , zn
zn +ℓ , ℓ ≥ 1ﻴﻤﻜﻥ ﺍﻥ ﺘﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜل
zn ( ℓ ) = ξ0 an + ξ1an −1 + ξ 2 an − 2 + ⋯ , ℓ ≥ 1
ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ zn +ℓ , ℓ ≥ 1ﺘﻜﺘﺏ ﺒﺩﻻﻝﺔ ﺍﻝﻨﻤﻭﺫﺝ ﻜﺎﻝﺘﺎﻝﻲ
zn + ℓ − µ = an + ℓ + ψ 1an + ℓ−1 + ⋯ + ψ ℓ−1an +1 + ψ ℓ an + ψ ℓ+1an −1 + ⋯ , ℓ ≥ 1
ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻝﺨﻁﺄ ﻴﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔ )ﺃﻨﻅﺭ ﺘﻌﺭﻴﻑ ( 5
E zn + ℓ − zn ( ℓ ) = E an + ℓ + ψ 1an +ℓ −1 + ⋯ + ψ ℓ−1an +1 + (ψ ℓ − ξ0 ) an + (ψ ℓ+1 − ξ1 ) an −1 + ⋯
2 2
∞
= (1 + ψ 12 + ⋯ + ψ ℓ2−1 ) σ 2 + ∑ (ψ ℓ+ j − ξ j ) σ 2
2
j =0
ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻝﺨﻁﺄ ﺍﻷﺩﻨﻰ ﻴﻨﺘﺞ ﻤﻥ ﺘﺼﻐﻴﺭ ﺍﻝﻌﻼﻗﺔ ﺍﻝﺴﺎﺒﻘﺔ ﺒﺎﻝﻨﺴﺒﺔ ﻝﻸﻭﺯﺍﻥ ξ jﻝﺠﻤﻴﻊ ﻗﻴﻡ j
ﻭﻫﺫﺍ ﻴﻤﻜﻥ ﺇﺫﺍ ﻭﻓﻘﻁ ﺇﺫﺍ ﺤﻘﻘﺕ ﺍﻷﻭﺯﺍﻥ ξ jﺍﻝﻌﻼﻗﺔ ﺍﻝﺘﺎﻝﻴﺔ
ξ j = ψ ℓ+ j , j = 0,1, 2,⋯ , ℓ ≥ 1
ﻭﻋﻠﻴﻪ ﻓﺈﻥ ﺍﻝﺘﻨﺒﺅﺍﺕ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻝﺨﻁﺄ ﺍﻷﺩﻨﻰ MMSE Forecastsﺘﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔ
zn ( ℓ ) = ψ ℓan + ψ ℓ +1an −1 + ψ ℓ+ 2 an −2 + ⋯ , ℓ ≥ 1
72
ﻨﻅﺭﻴﺔ :2
ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﺘﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔ:
en ( ℓ ) = zn + ℓ − zn ( ℓ ) = an +ℓ + ψ 1an + ℓ−1 + ψ 2 an +ℓ −2 + ⋯ + ψ ℓ−1an +1 , ℓ ≥ 1
ﻭﺘﺒﺎﻴﻥ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﺘﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔ:
V en ( ℓ ) = σ 2 (1 + ψ 12 + ψ 22 + ⋯ + ψ ℓ2−1 ) , ℓ ≥ 1
ﺍﻝﺼﻴﻐﺔ zn ( ℓ ) = ψ ℓan + ψ ℓ+1an −1 + ψ ℓ +2an −2 + ⋯, ℓ ≥ 1ﻏﻴﺭ ﻋﻤﻠﻴﺔ ﻹﻴﺠﺎﺩ ﺍﻝﺘﻨﺒﺅﺍﺕ ﻝﻠﻘﻴﻡ
ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ zn +ℓ , ℓ ≥ 1ﻭﺫﻝﻙ ﻷﻨﻨﺎ ﻨﺤﺘﺎﺝ ﺇﻝﻰ ﻤﻌﺭﻓﺔ ﺍﻝﻘﻴﻡ } . {a1 , a2 ,⋯ , an −1 , an
ﺘﻌﺭﻴﻑ : 18
ﺍﻝﻤﻌﻠﻭﻤﺎﺕ ﻤﺠﻤﻭﻋﺔ ﺘﻜﺎﻓﺊ )} I ({z1 , z2 ,⋯ , zn −1 , zn ﺍﻝﻤﻌﻠﻭﻤﺎﺕ ﻤﺠﻤﻭﻋﺔ
)} I ({a1 , a2 ,⋯, an−1 , anﻭﺫﻝﻙ ﺒﺎﻝﻤﻌﻨﻰ ﺃﻥ ﺍﻝﻤﺠﻤﻭﻋﺔ } {a1 , a2 ,⋯ , an −1 , anﺘﺤﺘﻭﻯ ﻋﻠﻰ ﻨﻔﺱ
ﺍﻝﻤﻌﻠﻭﻤﺎﺕ ﻋﻥ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ } . {z1 , z2 ,⋯ , zn−1 , zn
ﻤﻼﺤﻅﺔ :ﺍﻝﻤﺘﺴﻠﺴﺔ ﺍﻝﺯﻤﻨﻴﺔ } {z1 , z2 ,⋯, zn −1 , znﻴﻤﻜﻥ ﻤﺸﺎﻫﺩﺘﻬﺎ ﻭﻗﻴﺎﺴﻬﺎ ﻭﻝﻜﻥ ﺍﻝﻤﺘﻠﺴﻠﺔ
} {a1 , a2 ,⋯ , an −1 , anﻻﻴﻤﻜﻥ ﻤﺸﺎﻫﺩﺘﻬﺎ ﺃﻭ ﻗﻴﺎﺴﻬﺎ.
ﻨﻅﺭﻴﺔ : 3
ﺍﻝﻤﺘﻨﺒﺊ ﺫﺍ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻝﺨﻁﺄ ﺍﻷﺩﻨﻰ MMSE Forecastsﻴﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔ
zn ( ℓ ) = E ( zn + ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
ﺃﻱ ﻫﻭ ﺍﻝﺘﻭﻗﻊ ﺍﻝﺸﺭﻁﻲ ﻝﻠﻘﻴﻤﺔ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ zn +ℓ , ℓ ≥ 1ﻤﻌﻁﻰ } . {z1 , z2 ,⋯ , zn−1 , zn
ﺘﺴﺘﺨﺩﻡ ﻨﻅﺭﻴﺔ 2ﻋﻤﻠﻴﺎ ﻹﻴﺠﺎﺩ ﻗﻴﻡ ﺍﻝﺘﻨﺒﺅﺍﺕ ﺒﺩﻻ ﻤﻥ ﺍﻝﺼﻴﻐﺔ
zn ( ℓ ) = ψ ℓan + ψ ℓ +1an −1 + ψ ℓ+ 2 an −2 + ⋯ , ℓ ≥ 1
73
ﻭﺫﻝﻙ ﺘﺒﻌﺎ ﻝﻠﻤﻼﺤﻅﺔ ﺍﻝﺴﺎﺒﻘﺔ.
ﻗﺎﻋﺩﺓ :2
a , j ≤ 0
1 − E ( an + j zn , zn −1 ,⋯) = n + j
0, j>0
zn + j , j≤0
2 − E ( zn + j zn , zn −1 ,⋯) =
zn ( j ) , j > 0
ﻨﻅﺭﻴﺔ 3ﻤﻊ ﺍﻝﻘﺎﻋﺩﺓ 2ﺘﻌﻁﻲ ﻁﺭﻴﻘﺔ ﻋﻤﻠﻴﺔ ﻭﺴﻬﻠﺔ ﻹﻴﺠﺎﺩ ﺘﻨﺒﺅﺍﺕ ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ zn + ℓ , ℓ ≥ 1
ﺘﻌﺭﻴﻑ :19
ﺍﻝﺩﺍﻝﺔ zn ( ℓ ) , ℓ ≥ 1ﻜﺩﺍﻝﺔ ﻝﺯﻤﻥ ﺍﻝﺘﻘﺩﻡ ℓ ≥ 1ﻋﻨﺩ ﻨﻘﻁﺔ ﺍﻻﺼل ﻝﻠﺯﻤﻥ nﺘﺴﻤﻰ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ.
دوال ا 4ذج ): ARIMA(p,d,q
ﺍﻭﻻ :ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﻝﻨﻤﻭﺫﺝ ): AR(1
ﻝﻨﻔﺘﺭﺽ ﺍﻨﻨﺎ ﺸﺎﻫﺩﻨﺎ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ } {z1 , z2 ,⋯ , zn−1 , znﺤﺘﻰ ﺍﻝﺯﻤﻥ nﻭﺍﻝﺘﻲ ﻨﻌﺘﻘﺩ ﺍﻨﻬﺎ
ﺘﺘﺒﻊ ﻨﻤﻭﺫﺝ ) AR(1ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜل
) ∞ zt − µ = φ1 ( zt −1 − µ ) + at , at ∼ WN ( 0,σ 2 ) , φ1 < 1, µ ∈ ( −∞,
ﻨﺭﻴﺩ ﺃﻥ ﻨﺘﻨﺒﺄ ﻋﻥ ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ⋯ zn +1 , zn+2 , zn +3 ,ﺃﻭ ﺒﺸﻜل ﻋﺎﻡ . zn +ℓ , ℓ ≥ 1
ﻤﻥ ﻨﻅﺭﻴﺔ 3ﻨﺠﺩ
zn ( ℓ ) = E ( zn +ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
=µ +E φ1 ( zn + ℓ−1 − µ ) + an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
=µ +E φ1 ( zn + ℓ−1 − µ ) zn , zn −1 ,⋯ + an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
= µ +φ1E ( zn + ℓ−1 zn , zn −1 ,⋯) − µ + E an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
ﺃﻱ
74
zn ( ℓ ) = µ +φ1E ( zn + ℓ−1 zn , zn −1 ,⋯) − µ + E an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
ﻨﺤل ﻫﺫﻩ ﺍﻝﻌﻼﻗﺔ ﺘﻜﺭﺍﺭﻴﺎ ﻭﺒﺈﺴﺘﺨﺩﺍﻡ ﺍﻝﻘﺎﻋﺩﺓ 2
ℓ = 1: zn (1) = µ +φ1E ( zn zn , zn −1 ,⋯) − µ + E an +1 zn , zn −1 ,⋯
) = µ +φ1 ( zn − µ
ℓ = 2 : zn ( 2 ) = µ +φ1E ( zn +1 zn , zn −1 ,⋯) − µ + E an + 2 zn , zn −1 ,⋯
= µ +φ1 zn (1) − µ
ℓ = 3 : zn ( 3) = µ +φ1E ( zn + 2 zn , zn −1 ,⋯) − µ + E an +3 zn , zn −1 ,⋯
= µ +φ1 zn ( 2 ) − µ
ﻭﻫﻜﺫﺍ ﺒﺸﻜل ﻋﺎﻡ
zn ( ℓ ) = µ +φ1 zn ( ℓ − 1) − µ , ℓ ≥ 1
ﻭﻫﻲ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻷﺨﻁﺎﺀ ﺍﻷﺩﻨﻰ ﻝﻨﻤﻭﺫﺝ )AR(1
ﺘﻌﺭﻴﻑ :20
ﺸﺭﻁ ﺍﻹﺴﺘﻤﺭﺍﺭ Continuity Conditionﻴﺘﻁﻠﺏ ﺃﻨﻪ ﻋﻨﺩﻤﺎ ﺘﻜﻭﻥ ℓ = 1ﻓﺈﻥ
zn ( ℓ − 1) = zn ( 0 ) = zn
ﻤﻥ ﻨﻅﺭﻴﺔ 2ﺘﺒﺎﻴﻥ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﺘﻌﻁﻰ ﻤﻥ ﺍﻝﻌﻼﻗﺔ
V en ( ℓ ) = σ 2 (1 + ψ 12 + ψ 22 + ⋯ + ψ ℓ2−1 ) , ℓ ≥ 1
ﺴﺒﻕ ﺃﻥ ﺍﺸﺘﻘﻘﻨﺎ ﺩﺍﻝﺔ ﺍﻷﻭﺯﺍﻥ ﻝﻨﻤﻭﺫﺝ ) AR(1ﻭﻫﻲ
ψ j = φ1j , φ1 < 1
ﻭﺒﺎﻝﺘﻌﻭﻴﺽ ﻓﻲ ﺼﻴﻐﺔ ﺍﻝﺘﺒﺎﻴﻥ ﻨﺠﺩ
(
( V en ( ℓ ) = σ 2 1 + φ12 + φ14 + ⋯ + φ1
)2 ℓ −1
), ℓ ≥1
1 − φ12 ℓ
=σ2 , ℓ ≥1
1 − φ12
ﻤﺜﺎل :ﻤﺘﺴﻠﺴﻠﺔ ﺯﻤﻨﻴﺔ ﻤﺸﺎﻫﺩﻩ ﻭﺠﺩ ﺍﻨﻬﺎ ﺘﺘﺒﻊ ﺍﻝﻨﻤﻭﺫﺝ
75
) zt − 0.97 = 0.85 ( zt −1 − 0.97 ) + at , at ∼ WN ( 0, 0.024
ﺇﺫﺍ ﻜﺎﻨﺕ ﺍﻝﻤﺸﺎﻫﺩﺓ ﺍﻷﺨﻴﺭﺓ ﻫﻲ ، z156 = 0.49ﺃﻭﺠﺩ ﺘﻨﺒﺅﺍﺕ ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ z157 , z158 , z159
ﻭﺃﻭﺠﺩ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﻝﻬﺎ.
ﺍﻝﺤل :ﻤﻥ ﺍﻝﺼﻴﻐﺔ zn ( ℓ ) = µ +φ1 zn ( ℓ − 1) − µ , ℓ ≥ 1ﻨﺠﺩ
) z156 (1) = 0.97+0.85 ( z156 − 0.97
= 0.97+0.85 ( 0.49 − 0.97 ) = 0.56
) z156 ( 2 ) = 0.97+0.85 ( z156 (1) − 0.97
= 0.97+0.85 ( 0.56 − 0.97 ) = 0.62
) z156 ( 3) = 0.97+0.85 ( z156 ( 2 ) − 0.97
= 0.97+0.85 ( 0.62 − 0.97 ) = 0.68
ﻭﺍﻝﺘﺒﺎﻴﻨﺎﺕ
1 − φ12 ℓ
V en ( ℓ ) = σ 2 , ℓ ≥1
1 − φ12
V e156 (1) = 0.024
)1 − ( 0.85
4
V e156 ( 2 ) = 0.024 = 0.041
)1 − ( 0.85
2
)1 − ( 0.85
6
V e156 ( 2 ) = 0.024 = 0.054
)1 − ( 0.85
2
ﺜﺎﻨﻴﺎ :ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﻝﻨﻤﻭﺫﺝ ): AR(2
ﻝﻨﻔﺘﺭﺽ ﺍﻨﻨﺎ ﺸﺎﻫﺩﻨﺎ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ } {z1 , z2 ,⋯ , zn−1 , znﺤﺘﻰ ﺍﻝﺯﻤﻥ nﻭﺍﻝﺘﻲ ﻨﻌﺘﻘﺩ ﺍﻨﻬﺎ
ﺘﺘﺒﻊ ﻨﻤﻭﺫﺝ ) AR(2ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜل
zt = µ + φ1 ( zt −1 − µ ) + φ2 ( zt −2 − µ ) + at , at ∼ WN ( 0,σ 2 ) , µ ∈ ( −∞, ∞ ) ,
φ2 − φ1 < 1, φ2 + φ1 < 1, φ2 < 1
ﻨﺭﻴﺩ ﺃﻥ ﻨﺘﻨﺒﺄ ﻋﻥ ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ⋯ zn +1 , zn+2 , zn +3 ,ﺃﻭ ﺒﺸﻜل ﻋﺎﻡ . zn +ℓ , ℓ ≥ 1
ﻤﻥ ﻨﻅﺭﻴﺔ 3ﻨﺠﺩ
zn ( ℓ ) = E ( zn + ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
=µ +E φ1 ( zn + ℓ−1 − µ ) + φ2 ( zn + ℓ− 2 − µ ) + an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
76
=µ +E φ1 ( zn + ℓ−1 − µ ) zn , zn −1 ,⋯ + φ2 ( zn + ℓ−2 − µ ) zn , zn −1 ,⋯ + an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
= µ +φ1E ( zn + ℓ−1 zn , zn −1 ,⋯) − µ + φ2 E ( zn +ℓ −2 zn , zn −1 ,⋯) − µ + E an + ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
ﺃﻱ
zn ( ℓ ) = µ +φ1E ( zn + ℓ−1 zn , zn −1 ,⋯) − µ + φ2 E ( zn + ℓ−2 zn , zn −1 ,⋯) − µ + E an +ℓ zn , zn −1 ,⋯ , ℓ ≥ 1
2 ﻨﺤل ﻫﺫﻩ ﺍﻝﻌﻼﻗﺔ ﺘﻜﺭﺍﺭﻴﺎ ﻭﺒﺈﺴﺘﺨﺩﺍﻡ ﺍﻝﻘﺎﻋﺩﺓ
ℓ = 1: zn (1) = µ +φ1E ( zn zn , zn −1 ,⋯) − µ + φ2 E ( zn −1 zn , zn −1 ,⋯) − µ + E an +1 zn , zn −1 ,⋯
= µ +φ1 ( zn − µ ) + φ2 ( zn −1 − µ )
ℓ = 2 : zn ( 2 ) = µ +φ1E ( zn +1 zn , zn −1 ,⋯) − µ + φ2 E ( zn zn , zn −1 ,⋯) − µ + E an +2 zn , zn −1 ,⋯
= µ +φ1 zn (1) − µ + φ2 ( zn − µ )
ℓ = 3 : zn ( 3) = µ +φ1E ( zn + 2 zn , zn −1 ,⋯) − µ + φ2 E ( zn +1 zn , zn −1 ,⋯) − µ + E an +3 zn , zn −1 ,⋯
= µ +φ1 zn ( 2 ) − µ +φ2 zn (1) − µ
ℓ = 4 : zn ( 4 ) = µ +φ1E ( zn +3 zn , zn −1 ,⋯) − µ + φ2 E ( zn + 2 zn , zn −1 ,⋯) − µ + E an + 4 zn , zn −1 ,⋯
= µ +φ1 zn ( 3) − µ +φ2 zn ( 2 ) − µ
ﻭﻫﻜﺫﺍ ﺒﺸﻜل ﻋﺎﻡ
zn ( ℓ ) = µ +φ1 zn ( ℓ − 1) − µ + φ2 zn ( ℓ − 2 ) − µ , ℓ ≥ 1
AR(2) ﻭﻫﻲ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻷﺨﻁﺎﺀ ﺍﻷﺩﻨﻰ ﻝﻨﻤﻭﺫﺝ
ﻭﺩﺍﻝﺔ ﺍﻷﻭﺯﺍﻥ2 ﻭﻴﻤﻜﻥ ﺤﺴﺎﺏ ﺘﺒﺎﻴﻨﺎﺕ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﻤﻥ ﻨﻅﺭﻴﺔ
1, j=0
φ , j =1
1
ψj = 2
φ1 + φ2 , j=2
φψ
1 j −1 + φ2ψ j − 2 , j≥3
: ARIMA(0,1,1) ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﻝﻨﻤﻭﺫﺝ:ﺜﺎﻝﺜﺎ
ﻭﺍﻝﺘﻲ ﻨﻌﺘﻘﺩ ﺍﻨﻬﺎn { ﺤﺘﻰ ﺍﻝﺯﻤﻥz1 , z2 ,⋯ , zn−1 , zn } ﻝﻨﻔﺘﺭﺽ ﺍﻨﻨﺎ ﺸﺎﻫﺩﻨﺎ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ
ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜلARIMA(0,1,1) ﺘﺘﺒﻊ ﻨﻤﻭﺫﺝ
zt = zt −1 + at − θ1at −1 , at ∼ WN ( 0, σ 2 )
77
ﻨﺭﻴﺩ ﺃﻥ ﻨﺘﻨﺒﺄ ﻋﻥ ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ⋯ zn +1 , zn+2 , zn +3 ,ﺃﻭ ﺒﺸﻜل ﻋﺎﻡ . zn +ℓ , ℓ ≥ 1
ﻤﻥ ﻨﻅﺭﻴﺔ 3ﻨﺠﺩ
zn ( ℓ ) = E ( zn + ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
=E ( zn + ℓ−1 zn , zn −1 ,⋯) + E ( an + ℓ zn , zn −1 ,⋯) − θ1E ( an + ℓ−1 zn , zn −1 ,⋯) , ℓ ≥ 1
ﻨﺤل ﻫﺫﻩ ﺍﻝﻌﻼﻗﺔ ﺘﻜﺭﺍﺭﻴﺎ ﻭﺒﺈﺴﺘﺨﺩﺍﻡ ﺍﻝﻘﺎﻋﺩﺓ 2
zn ( ℓ ) =E ( zn + ℓ−1 zn , zn −1 ,⋯) + E ( an +ℓ zn , zn −1 ,⋯) − θ1E ( an + ℓ−1 zn , zn −1 ,⋯) , ℓ ≥ 1
)⋯ℓ = 1: zn (1) =E ( zn zn , zn −1 ,⋯) + E ( an +1 zn , zn −1 ,⋯) − θ1 E ( an zn , zn −1 ,
= zn − θ1an
)⋯ℓ = 2 : zn ( 2 ) =E ( zn +1 zn , zn −1 ,⋯) + E ( an +2 zn , zn −1 ,⋯) − θ1 E ( an +1 zn , zn −1 ,
)= zn (1
)⋯ℓ = 3 : zn ( 3) =E ( zn +2 zn , zn −1 ,⋯) + E ( an +3 zn , zn −1 ,⋯) − θ1E ( an +1 zn , zn −1 ,
) = zn ( 2
ﻭﻫﻜﺫﺍ ﺒﺸﻜل ﻋﺎﻡ
zn ( ℓ ) = zn ( ℓ − 1) , ℓ ≥ 2
ﻭﻫﻜﺫﺍ ﻓﺈﻥ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻷﺨﻁﺎﺀ ﺍﻷﺩﻨﻰ ﻝﻨﻤﻭﺫﺝ ) ARIMA(0,1,1ﺘﻌﻁﻰ
ﺒﺎﻝﻌﻼﻗﺔ
zn − θ1an , ℓ = 1
zn ( ℓ ) =
zn ( ℓ − 1) , ℓ > 1
ﺭﺍﺒﻌﺎ :ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﻝﻨﻤﻭﺫﺝ ): MA(1
ﻝﻨﻔﺘﺭﺽ ﺍﻨﻨﺎ ﺸﺎﻫﺩﻨﺎ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ } {z1 , z2 ,⋯ , zn−1 , znﺤﺘﻰ ﺍﻝﺯﻤﻥ nﻭﺍﻝﺘﻲ ﻨﻌﺘﻘﺩ ﺍﻨﻬﺎ
ﺘﺘﺒﻊ ﻨﻤﻭﺫﺝ ) MA(1ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜل
) zt = µ + at − θ1at −1 , at ∼ WN ( 0, σ 2
ﻨﺭﻴﺩ ﺃﻥ ﻨﺘﻨﺒﺄ ﻋﻥ ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ⋯ zn +1 , zn+2 , zn +3 ,ﺃﻭ ﺒﺸﻜل ﻋﺎﻡ . zn +ℓ , ℓ ≥ 1
ﻤﻥ ﻨﻅﺭﻴﺔ 3ﻨﺠﺩ
zn ( ℓ ) = E ( zn + ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
=µ + E ( an + ℓ zn , zn −1 ,⋯) − θ1 E ( an + ℓ−1 zn , zn −1 ,⋯) , ℓ ≥ 1
ﻨﺤل ﻫﺫﻩ ﺍﻝﻌﻼﻗﺔ ﺘﻜﺭﺍﺭﻴﺎ ﻭﺒﺈﺴﺘﺨﺩﺍﻡ ﺍﻝﻘﺎﻋﺩﺓ 2
78
zn ( ℓ ) =µ + E ( an + ℓ zn , zn −1 ,⋯) − θ1E ( an + ℓ−1 zn , zn −1 ,⋯) , ℓ ≥ 1
ℓ = 1: zn (1) =µ + E ( an +1 zn , zn −1 ,⋯) − θ1 E ( an zn , zn −1 ,⋯)
= µ − θ1an
ℓ = 2 : zn ( 2 ) =µ + E ( an + 2 zn , zn −1 ,⋯) − θ1 E ( an +1 zn , zn −1 ,⋯)
=µ
ℓ = 3 : zn ( 3) =µ + E ( an +3 zn , zn −1 ,⋯) − θ1 E ( an + 2 zn , zn −1 ,⋯)
=µ
ﻭﻫﻜﺫﺍ ﺒﺸﻜل ﻋﺎﻡ
zn ( ℓ ) = µ , ℓ ≥ 2
ﺘﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔMA(1) ﻭﻫﻜﺫﺍ ﻓﺈﻥ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻷﺨﻁﺎﺀ ﺍﻷﺩﻨﻰ ﻝﻨﻤﻭﺫﺝ
µ − θ1an , ℓ = 1
zn ( ℓ ) =
µ, ℓ≥2
: MA(2) ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﻝﻨﻤﻭﺫﺝ: ﺨﺎﻤﺴﺎ
ﻭﺍﻝﺘﻲ ﻨﻌﺘﻘﺩ ﺍﻨﻬﺎn { ﺤﺘﻰ ﺍﻝﺯﻤﻥz1 , z2 ,⋯ , zn−1 , zn } ﻝﻨﻔﺘﺭﺽ ﺍﻨﻨﺎ ﺸﺎﻫﺩﻨﺎ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ
ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜلMA(2) ﺘﺘﺒﻊ ﻨﻤﻭﺫﺝ
zt = µ + at − θ1at −1 − θ 2 at −2 , at ∼ WN ( 0,σ 2 )
. zn +ℓ , ℓ ≥ 1 ﺃﻭ ﺒﺸﻜل ﻋﺎﻡzn +1 , zn+2 , zn +3 ,⋯ ﻨﺭﻴﺩ ﺃﻥ ﻨﺘﻨﺒﺄ ﻋﻥ ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ
ﻨﺠﺩ3 ﻤﻥ ﻨﻅﺭﻴﺔ
zn ( ℓ ) = E ( zn +ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
=µ + E ( an + ℓ zn , zn −1 ,⋯) − θ1 E ( an +ℓ −1 zn , zn −1 ,⋯) − θ 2 E ( an +ℓ −2 zn , zn −1 ,⋯) , ℓ ≥ 1
2 ﻨﺤل ﻫﺫﻩ ﺍﻝﻌﻼﻗﺔ ﺘﻜﺭﺍﺭﻴﺎ ﻭﺒﺈﺴﺘﺨﺩﺍﻡ ﺍﻝﻘﺎﻋﺩﺓ
zn ( ℓ ) =µ + E ( an + ℓ zn , zn −1 ,⋯) − θ1 E ( an + ℓ−1 zn , zn −1 ,⋯) − θ 2 E ( an + ℓ−2 zn , zn −1 ,⋯) , ℓ ≥ 1
ℓ = 1: zn (1) =µ + E ( an +1 zn , zn −1 ,⋯) − θ1 E ( an zn , zn −1 ,⋯) − θ 2 E ( an −1 zn , zn −1 ,⋯)
= µ − θ1an − θ 2 an −1
ℓ = 2 : zn ( 2 ) =µ + E ( an +2 zn , zn −1 ,⋯) − θ1E ( an +1 zn , zn −1 ,⋯) − θ 2 E ( an zn , zn −1 ,⋯)
= µ − θ 2 an
ℓ = 3 : zn ( 3) =µ + E ( an +3 zn , zn −1 ,⋯) − θ1E ( an + 2 zn , zn −1 ,⋯) − θ 2 E ( an +1 zn , zn −1 ,⋯)
=µ
ﻭﻫﻜﺫﺍ ﺒﺸﻜل ﻋﺎﻡ
79
zn ( ℓ ) = µ , ℓ ≥ 3
ﺘﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔMA(2) ﻭﻫﻜﺫﺍ ﻓﺈﻥ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻷﺨﻁﺎﺀ ﺍﻷﺩﻨﻰ ﻝﻨﻤﻭﺫﺝ
µ − θ1an − θ 2 an −1 , ℓ = 1
z n ( ℓ ) = µ − θ 2 an , ℓ=2
µ, ℓ≥3
: ARMA(1,1) ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﻝﻨﻤﻭﺫﺝ: ﺴﺎﺩﺴﺎ
ﻭﺍﻝﺘﻲ ﻨﻌﺘﻘﺩ ﺍﻨﻬﺎn { ﺤﺘﻰ ﺍﻝﺯﻤﻥz1 , z2 ,⋯ , zn−1 , zn } ﻝﻨﻔﺘﺭﺽ ﺍﻨﻨﺎ ﺸﺎﻫﺩﻨﺎ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ
ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜلARMA(1,1) ﺘﺘﺒﻊ ﻨﻤﻭﺫﺝ
zt = µ + φ1 ( zt −1 − µ ) + at − θ1at −1 , at ∼ WN ( 0,σ 2 ) , φ1 ≠ θ1 , φ1 < 1
. zn +ℓ , ℓ ≥ 1 ﺃﻭ ﺒﺸﻜل ﻋﺎﻡzn +1 , zn+2 , zn +3 ,⋯ ﻨﺭﻴﺩ ﺃﻥ ﻨﺘﻨﺒﺄ ﻋﻥ ﺍﻝﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ
ﻨﺠﺩ3 ﻤﻥ ﻨﻅﺭﻴﺔ
zn ( ℓ ) = E ( zn +ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
=µ + φ1 E ( zn + ℓ−1 − µ ) zn , zn −1 ,⋯ + E ( an +ℓ zn , zn −1 ,⋯) − θ1E ( an + ℓ−1 zn , zn −1 ,⋯) , ℓ ≥ 1
2 ﻨﺤل ﻫﺫﻩ ﺍﻝﻌﻼﻗﺔ ﺘﻜﺭﺍﺭﻴﺎ ﻭﺒﺈﺴﺘﺨﺩﺍﻡ ﺍﻝﻘﺎﻋﺩﺓ
zn ( ℓ ) =µ + φ1 E ( zn + ℓ−1 − µ ) zn , zn −1 ,⋯ + E ( an + ℓ zn , zn −1 ,⋯) − θ1 E ( an +ℓ −1 zn , zn −1 ,⋯) , ℓ ≥ 1
ℓ = 1: zn (1) =µ + φ1 E ( zn − µ ) zn , zn −1 ,⋯ + E ( an +1 zn , zn −1 ,⋯) − θ1E ( an zn , zn −1 ,⋯)
= µ + φ1 ( zn − µ ) − θ1an
ℓ = 2 : zn ( 2 ) =µ + φ1E ( zn +1 − µ ) zn , zn −1 ,⋯ + E ( an + 2 zn , zn −1 ,⋯) − θ1 E ( an +1 zn , zn −1 ,⋯)
= µ + φ1 zn (1) − µ
ℓ = 3 : zn ( 3) =µ + φ1 E ( zn + 2 − µ ) zn , zn −1 ,⋯ + E ( an +3 zn , zn −1 ,⋯) − θ1E ( an + 2 zn , zn −1 ,⋯)
= µ + φ1 zn ( 2 ) − µ
ﻭﻫﻜﺫﺍ ﺒﺸﻜل ﻋﺎﻡ
zn ( ℓ ) = µ + φ1 zn ( ℓ − 1) − µ , ℓ ≥ 2
ﺘﻌﻁﻰARMA(1,1) ﻭﻫﻜﺫﺍ ﻓﺈﻥ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺫﺍﺕ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﻷﺨﻁﺎﺀ ﺍﻷﺩﻨﻰ ﻝﻨﻤﻭﺫﺝ
ﺒﺎﻝﻌﻼﻗﺔ
µ + φ1 ( zn − µ ) − θ1an , ℓ =1
zn ( ℓ ) =
µ + φ1 zn ( ℓ − 1) − µ , ℓ≥2
80
ﺘﻤﺭﻴﻥ:
ﻝﻨﻤﻭﺫﺝ ) ARMA(1,1ﻭﺍﻝﺫﻱ ﻴﻜﺘﺏ ﻋﻠﻰ ﺍﻝﺸﻜل
zt = µ + φ1 ( zt −1 − µ ) + at − θ1at −1 , at ∼ WN ( 0,σ 2 ) , φ1 ≠ θ1 , φ1 < 1
ﺒﺭﻫﻥ ﺍﻥ ﻋﻨﺩﻤﺎ ﺘﺅﻭل φ1 → 1ﻓﺈﻥ ) ARMA(1,1) → IMA(1,1ﻭﻤﻥ ﺜﻡ ﺃﻭﺠﺩ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ
ﻝﻨﻤﻭﺫﺝ ). IMA(1,1
ﺘﻤﺭﻴﻥ:
ﺃﻭﺠﺩ ﺩﻭﺍل ﺍﻝﺘﻨﺒﺅ ﻭﺘﺒﺎﻴﻥ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﻝﻜل ﻤﻥ ﺍﻝﻨﻤﺎﺫﺝ ﺍﻝﺘﺎﻝﻴﺔ:
ARIMA(1,1,1), ARIMA(2,1,0), ARIMA(0,1,2), ARIMA(1,2,0),
ARIMA(0,2,1), ARIMA(0,2,0).
ﺤﺩﻭﺩ ﺍﻝﺘﻨﺒﺅ : Forecasting Limits
ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ zn ( ℓ ) , ℓ ≥ 1ﻋﻨﺩ ﻗﻴﻤﺔ ﻤﻌﻴﻨﺔ ﺘﻌﻁﻲ ﻤﺎﻴﺴﻤﻰ ﺒﺘﻨﺒﺅ ﺍﻝﻨﻘﻁﺔ Point Forecastﻭﺍﻝﺫﻱ
ﻻﻴﻜﻔﻲ ﺍﻭ ﻴﻔﻴﺩ ﻓﻲ ﺇﺘﺨﺎﺫ ﻗﺭﺍﺭﺍﺕ ﺇﺤﺼﺎﺌﻴﺔ ﻋﻥ ﺍﻝﻅﺎﻫﺭﺓ ﺍﻝﻌﺸﻭﺍﺌﻴﺔ ﺍﻝﻤﺩﺭﻭﺴﺔ ﻷﻥ
P ( Z n + m = zn ( m ) ) = 0, for some m > 0
ﺃﻱ ﺃﻥ ﻤﻘﺩﺍﺭ ﺘﺄﻜﺩﻨﺎ ) ﺃﻭ ﺇﺤﺘﻤﺎل( ﻤﻥ ﺃﻥ ﺍﻝﻘﻴﻤﺔ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ﺍﻝﻤﺭﺍﺩ ﺍﻝﺘﻨﺒﺅ ﻋﻨﻬﺎ ﺘﺴﺎﻭﻱ ﺍﻝﻘﻴﻤﺔ
ﺍﻝﻤﻌﻁﺎﺓ ﻤﻥ ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺘﺴﺎﻭﻱ ﺍﻝﺼﻔﺭ ﺃﻱ ﺍﻨﻨﺎ ﻏﻴﺭ ﻤﺘﺄﻜﺩﻴﻥ ﺇﻁﻼﻗﺎ ﻭﺒﺎﻝﺘﺎﻝﻲ ﻻﻓﺎﺌﺩﺓ ﻤﻥ ﺍﻝﺘﻨﺒﺅ.
ﻝﻠﺘﻐﻠﺏ ﻋﻠﻰ ﺫﻝﻙ ﻭﺃﻹﺴﺘﻔﺎﺩﺓ ﻤﻥ ﺍﻝﺘﻨﺒﺅﺍﺕ ﻨﺴﺘﺨﺩﻡ ﻤﺎﻴﺴﻤﻰ ﺒﺘﻨﺒﺅ ﺍﻝﻔﺘﺭﺓ Interval Forecast
ﻭﻫﻲ ﻋﺒﺎﺭﺓ ﻋﻥ ﻓﺘﺭﺓ ﻤﺜل ] [a, bﻋﻠﻰ ﺨﻁ ﺍﻷﻋﺩﺍﺩ ﺍﻝﺤﻘﻴﻘﻴﺔ ﺒﺤﻴﺙ ﻴﻜﻭﻥ
) P ( a ≤ Z n + m ≤ b ) = (1 − α
ﻭﺒﻬﺫﺍ ﻨﺴﺘﻁﻴﻊ ﺃﻥ ﻨﺤﺩﺩ ﺩﺭﺠﺔ ﺘﺄﻜﺩﻨﺎ ﻤﻥ ﺃﻥ ﺍﻝﻘﻴﻤﺔ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ﺍﻝﻤﺭﺍﺩ ﺍﻝﺘﻨﺒﺅ ﻋﻨﻬﺎ ﺘﻘﻊ ﺒﻴﻥ ﺍﻝﻘﻴﻡ a
ﻭ bﺒﺩﺭﺠﺔ ﺘﺄﻜﺩ ﺃﻭ ﺇﺤﺘﻤﺎل ) 1 − αﺃﻭ ( 100 × (1 − α ) %ﻓﻤﺜﻼ ﻝﻭ ﻜﺎﻨﺕ α = 0.05ﻓﺈﻨﻨﺎ
ﻨﻜﻭﻥ ﻤﺘﺄﻜﺩﻴﻥ ﻭﺒﺈﺤﺘﻤﺎل 95%ﺍﻥ ﺍﻝﻘﻴﻤﺔ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ ﺘﻘﻊ ﺒﻴﻥ ﺍﻝﻘﻴﻡ aﻭ . b
ﺘﻌﺭﻴﻑ :21
ﻋﻠﻰ ﺇﻓﺘﺭﺍﺽ ﺃﻥ ) at ∼ N ( 0, σ 2ﻓﺈﻥ ﺤﺩﻭﺩ 100 × (1 − α ) %ﻓﺘﺭﺓ ﺘﻨﺒﺅ ﻝﻠﻘﻴﻤﺔ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ
zn +ℓ , ℓ ≥ 1ﺘﻌﻁﻰ ﺒﺎﻝﻌﻼﻗﺔ
81
{ }
zn ( ℓ ) ± uα 2 V en ( ℓ )
12
α
ﺤﻴﺙ uα 2ﺍﻝﻤﺌﻴﻥ 100 1 − ﻝﻠﺘﻭﺯﻴﻊ ). N ( 0,1
2
ﻓﻤﺜﻼ ﻋﻨﺩﻤﺎ α = 0.05ﻓﺈﻥ . u0.025 = 1.96
ﻤﻼﺤﻅﺔ :ﻓﻲ ﺍﻝﺘﻌﺭﻴﻑ ﺇﻓﺘﺭﻀﻨﺎ ﺃﻥ ﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﻀﺠﺔ ﺍﻝﺒﻴﻀﺎﺀ ) at ∼ N ( 0, σ 2ﻭﻫﺫﺍ ﻤﻤﻜﻥ
ﺇﻋﺘﻤﺎﺩﺍ ﻋﻠﻰ ﻨﻅﺭﻴﺔ ﻨﻬﺎﻴﺔ ﻤﺭﻜﺯﻴﺔ.
ﻤﺜﺎل :ﻤﺘﺴﻠﺴﻠﺔ ﺯﻤﻨﻴﺔ ﻤﺸﺎﻫﺩﻩ ﻭﺠﺩ ﺍﻨﻬﺎ ﺘﺘﺒﻊ ﺍﻝﻨﻤﻭﺫﺝ
) zt − 0.97 = 0.85 ( zt −1 − 0.97 ) + at , at ∼ N ( 0,0.024
ﺇﺫﺍ ﻜﺎﻨﺕ ﺍﻝﻤﺸﺎﻫﺩﺓ ﺍﻷﺨﻴﺭﺓ ﻫﻲ ، z156 = 0.49ﺃﻭﺠﺩ ﺘﻨﺒﺅﺍﺕ ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ z157 , z158 , z159
ﻭﺃﻭﺠﺩ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﻝﻬﺎ ﻭﻤﻥ ﺜﻡ ﺃﻭﺠﺩ ﻓﺘﺭﺍﺕ ﺘﻨﺒﺅ 95%ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ.
ﺍﻝﺤل :ﺴﺒﻕ ﺃﻥ ﺤﺴﺒﻨﺎ ﻓﻲ ﻤﺜﺎل ﺴﺎﺒﻕ ﺍﻝﺘﻨﺒﺅﺍﺕ ﻭ ﺃﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﻜﺎﻝﺘﺎﻝﻲ:
ﺍﻝﺘﻨﺒﺅﺍﺕ
z156 (1) = 0.56, z156 ( 2 ) = 0.62, z156 ( 3) = 0.68
ﻭﺍﻝﺘﺒﺎﻴﻨﺎﺕ
V e156 (1) = 0.024, V e156 ( 2 ) = 0.041, V e156 ( 2 ) = 0.054
ﻓﺘﺭﺍﺕ ﺘﻨﺒﺅ 95%ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ z157 , z158 , z159ﻨﻭﺠﺩﻫﺎ ﻤﻥ ﺼﻴﻐﺔ ﺘﻌﺭﻴﻑ 21
{
zn ( ℓ ) ± uα 2 V en ( ℓ ) }
12
{
1 − z156 (1) ± u0.025 V e156 (1) }
12
= 0.56 ± 1.96 0.024 = 0.56 ± 0.304
2 − z156 ( 2 ) ± u0.025 {V e }( 2 )
12
156 = 0.62 ± 1.96 0.041 = 0.62 ± 0.397
3 − z156 ( 3) ± u0.025 {V e }( 3)
12
156 = 0.68 ± 1.96 0.054 = 0.68 ± 0.455
ﺃﻱ ﺃﻥ ) z157 ∈ ( 0.256,0.864ﺒﺈﺤﺘﻤﺎل 0.95ﻭ ﻭﻜﺫﻝﻙ ) z158 ∈ ( 0.223,1.017ﻭ ﻜﺫﻝﻙ ﺃﻴﻀﺎ
). z159 ∈ ( 0.225,1.135
82
ﺍﻟﻔﺼﻞ ﺍﻟﺨﺎﻣﺲ
"Mو$ء !7م 4إDesigning and Building Statistical MK
: Forecasting System
( أن ذآ ان ا
[16ة او
; !) Iم :('Cه'. 7ء !1ذج .إن !'. 4ء !1ذج
إ 78 5ه3C 4! 7ار13;C Iterative 42ن & 2%Cا
'!1ذج 2JC ،ا
'!1ذج )و O. J
)
"& 2JCا
'!1ذج( و إ;(ر ا
'!1ذج.
#أو /ا ذج ): Model Identification (Specification
2%C 45& 7Nا
'!1ذج ;[م ا
( ت أو ا
!Kهات ا
) 4J.ا
;ر (r2واي &"&1ت اى
ا
4b3ا
;
1C 7ت O.ا
!; 4وذ
;B9 fاح & ! & 41ا
'!ذج ا
!'( .4و "C );2أو
2%Cا
'!1ذج W 5ا
[16ات ا
" 4`2ا
;
:4
ا ;:ة اrو' S- . / :ا: Variance-stabilizing Transformation #
".ر) ا
!; w6[& 7N 4ز&' Time Plot 7وإLاء \".ا(;9رات ا4N"!
48 59
!Nإذا آن ا
;( ،S.U 2و 4
5 7Nم (Uت ا
;( 2او إذا آن ا
;(1; & V& Y;2 2ى
ا
!; (6 ' FN 4ا
; H21%ا
z1ر 7!Cا
!; 4و FN 2L & O %bذا S(]C )C
ا
;( 2وإ { Xإ
(6C 7أ 5ا
; 21%ت ا
; 7ذآ ه L 7Nول .41 4%b+
ا ;:ة ا : 7-إ?ر در nا: d = Y
إذا آ Sا
!; 4أو J; & z O21%Cة 7Nا
!; 2%C ' W N w1در 4Lا
; d 2bا
;7
H" Cا
!!; 4أو J; & O21%Cة 7Nا
!; w1و 1Jم A{.ا
; 2bاول o%b )Uا
;
:7
-1ا
![66ت ا
&' 4 ;!!
4أو .O21%C
66[& -2ت دا
; 7ا
;ا w.ا
Aا 7Cا
"' 7وا
;ا w.ا
Aا 7Cا
78ا
"' SACF 7و . SPACF
-3إLاء , 2bCإذا ا;5ج ا& وإدة ا
[16ات 1و 2ا
. ;J.
ا
![66ت ا
&' ;!
4ت zا
! ;Jة 7N YC (Cا
! ;1ى ودا
C 4ا w.ذا7' 7C
&;[&ة w(.ء آ! ان دا
4ا
;ا w.ا
Aا 7Cا
78ا
"' 4!B 76"C 7وا5ة & 4(2Bا
1ا5
ا
\Y.) ^%ا
' Iا9رة( و 4J.ا
L 4(2B )Jا & ا
.b
& :4I5در 4Lا
;13C & (
z d 2bن 0او 1او . 2
83
: q وp / : --; ة ا:ا
ار%
9 ا4L در2%;. م1J
w1;!
وا2(;
& اH آ7N ةJ; & 4 ;& H %
" ان.
w. وا
;ا7'"
ا7CاA
اw. ا
;ا7;
أ
!ط دا4
رJ!. f
وذq ك%;!
اw1;!
ا4L ودرp 7CاA
ا
78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
ا42I'
ا
!ط اV& 7'"
ا78
ا7CاA
ا
@A ه76"2 7
;
وا
ول ا38 4%b+ 7N رة1آA!
اARMA(p,q) اص
!ذج1[. 2; &
:4"8K
اص
("\ ا
'!ذج ا1[
ا
PACF ACF ذج1!'
ا
φkk = 0, k > 1 &;دد7 أوا7[& اC AR(1)( و1,d,0)
φkk = 0, k > 2 7(L &[C او7[& اC AR(2)( و2,d,0)
φkk = 0, k > p 7(L &[C او/ و7[& اC AR(p)( وp,d,0)
7[& اC O 6 2 ρ k = 0, k > 1 MA(1) ( و0,d,1)
7(L او7[& اC O 6 2 ρ k = 0, k > 2 MA(2) ( و0,d,2)
7(L او/ و7[& اC O 6 2 ρ k = 0, k > q MA(q) ( و0,d,q)
7[& اC O 6 2 وoB';C ;[& اC وoB';C ARMA(1,1) ( و1,d,1)
1 e[;
& ا 1 e[;
& ا
e[;
" ا. oB';C &[;C وq - p e[;
" ا. oB';C ARMA(p,q) ( وp,d,q)
6 2 وp – q q – p e[;
" ا. (L او/ ا و
7[& اC O
". 7(L او/و
p – q e[;
ا
:
افV7
" إaA إ: $; ة ا
ا:ا
م1"& إ
اف4NP! إذا آن ' إN ' ا
;{آW N 2bC
;ج إ%C 4 ;!
اS
إذا آ
{6[
اV& ةJ; !
ا4Bb!
ا4 ;!
w 4'"
اw1;& 4
رJ!. );2 اAذج وه1!'
إ
اδ
w1;!
ا اAO
ا
!"ري
12
c
s.e ( w ) ≅ 0 (1 + 2r1 + 2 r2 + ⋯ + 2 rK )
n
84
;(ر9ن ا132 و. K 4L
ر421'"!
ا4'"
ا4CاA
ت ا6. ا
;ا7 هr1 ,⋯ , rK وc0 = γˆ0 Q5
H0 : δ = 0
H1 : δ ± 0
w
. > 1.96 S
إذاآα = 0.05 ' H 0 \N
و
s.e ( w )
85
: Model Estimation
ا ذج
f
وذσ 2 وθ1 ,… ,θ q وφ1 ,… ,φ p وδ ذج1!'
&"
) ا2JC & .X ذج1!'
اH3 2%C ".
.'2
ةN1;!
ا4[2;[ام ا
(
ت ا
;رF.
;حJ!
ذج ا1!'
واz1 , z2 ,… , zn −1 , zn هةK!
ا4'&
ا4 ;!
' ا2
;ض انb'
φ p ( B ) wt = δ + θ q ( B ) at , at ∼ N ( 0, σ 2 )
أو
φ p ( B ) zt = δ + θ q ( B ) at , at ∼ N ( 0, σ 2 )
O"!L VJC φ p ( B ) = 0 4
ور ا
!"دAL و4;آK& ورAL O'. L12X θ q ( B ) وφ p ( B ) Q5
.(ارJ;9ة ) ط ا51
ة ا8رج دا
رJ!
ا اAق ه6
!P HC ;J2= wJN ' هO'& آA' )
"!
ا2J;
ه'ك =ق آ]ة
.4=K
"ت ا
ا.!
ا4J2= ا
"وم و4J2= !وه
: The Method of Moments
اومi :rأو
"وم. rk 4'"
4CاA
ت ا6. وا
;اz 4'"
اw1;& H]& 4'"
";! & وات وم اCو
!"
) ا
!اد4( '
. 4 C'
ت اX ا
!"دH5 وρ k 7CاA
اw. ا
;ا4
وداµ w1;!
اH]& 42I'
ا
.ه2JC
:7
;
آAR(p) ذج1!'
4J26
ف
;"ض ا1
µˆ = z = ∑ i =1 zi n ايz رJ!
. µ w1;!
ر اJ2 -1
n
:4B "
;[م اφ1 ,…, φ p 2J;
-2
ρ k = φ1 ρ k −1 + φ2 ρ k −2 + ⋯ + φ p ρ k − p , k > 1
7N .VB1;
اA وأzt −k − µ %
. AR(p) ذج1!'
4N"!
ا4
ب ا
!"دP & _;'C 7;
وا
ل و12 تXت ا
! ! &"دXم ا
!"دI
H %
k = 1,2,…, p VP1. 4J.
ا4
ا
!"د
:7
;
اYule-Walker ووآ
ρ1 = φ1 + φ2 ρ1 + ⋯ + φ p ρ p −1
ρ 2 = φ1 ρ1 + φ2 + ⋯ + φ p ρ p −2
⋮
ρ p = φ1 ρ p −1 + φ2 ρ p −2 + ⋯ + φ p
:7
;
آφˆ1 ,… ,φˆp )
"!
رات ا
"ومJ& H %
rk رJ!
. ρ k \21";
. و
86
:7N1b !
اH3K
ل و ووآ ا12 تX &"دVP1.
r1 1 r1 r2 ⋯ rp −2 rp −1 φˆ1
r r
1 r1 ⋯ rp −3 rp −2 φˆ2
2= 1
⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
rp rp −1 rp −2 rp −3 ⋯ r1 1 φˆ
p
)
"!
4
@ ا
!"دA هH%.و
φˆ1 1 r1 r2 ⋯ rp −2 rp −1
−1
r1
φˆ2 r1 1 r1 ⋯ rp −3 rp −2 r
2
⋮ = ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
φˆ rp −1 rp −2 rp −3 ⋯ r1 1 rp
p
7
;
آσ 2 رJC
(
σˆ 2 = γˆ0 1 − φˆ1r1 − φˆ2 r2 −⋯φˆp rp )
Q5
1 n
γˆ0 = ∑ ( zt − z )
2
n t =1
.4'"
ا2(C 1ه
: اذجv
اوم
AR(1) ذج7 -1
zt − µ = φ1 ( zt −1 − µ ) + at , at ∼ N ( 0,σ 2 )
1 هφ1 )"!
ر ا
"ومJ&
φˆ1 = r1
1 هµ )"!
ر ا
"ومJ&
µˆ = z
1 هσ 2 )"!
ر ا
"ومJ&
(
σˆ 2 = γˆ0 1 − φˆ1r1 )
Q5
1 n
γˆ0 = ∑ ( zt − z )
2
n t =1
87
MA(1) ذج7 -2
zt − µ = at − θ1at −1 , at ∼ N ( 0,σ 2 )
4B "
;[م اθ1 )"!
ر ا
"ومJ& د29
−θ1
ρ1 =
1 + θ12
OCراJ!. )
"!
\ ا21";.و
−θˆ1
r1 =
1 + θˆ12
θˆ1 رJ!
4
ا
!"دH%.و
−1 ± 1 − 4 r1
θˆ1 =
2 r1
نFN r1 = −0.4 S آ1
]!N . θˆ1 < 1 J%C 7;
ا4!J
اA{ θˆ1 رJ!
;!B 76"2 H%
ا اAه
. θˆ1 = −0.77 1 هθ1 )"!
ر ا
"ومJ& ن132 7
;
.( وθˆ1 ) = 3.27 ( وθˆ1 ) = −0.77
2 1
AR(2) ذج7 -3
zt − µ = φ1 ( zt −1 − µ ) + φ2 ( zt −2 − µ ) + at , at ∼ N ( 0,σ 2 )
7 هφ2 وφ1 )
"!
رات ا
"ومJ& ل ووآ12 تX;[ام &"دF.
φˆ1 1 r1 −1 r1
=
φˆ r1 1 r2
2
O'&و
r1 − r1r2 ˆ = r2 − r1
2
φˆ1 = , φ
1 − r12 1 − r12
2
1 هµ )"!
ر ا
"ومJ&
µˆ = z
1 هσ 2 )"!
ر ا
"ومJ&
(
σˆ 2 = γˆ0 1 − φˆ1r1 − φˆ2 r2 )
Q5
1 n
γˆ0 = ∑ ( zt − z )
2
n t =1
88
MA(2) ذج7 -4
zt − µ = at − θ1at −1 − θ 2 at −2 , at ∼ N ( 0, σ 2 )
تB "
;[م اθ 2 وθ1 )
"!
رات ا
"ومJ& د29
−θ1 (1 − θ 2 ) −θ 2
ρ1 = , ρ2 =
1 + θ1 + θ 2
2 2
1 + θ12 + θ 22
θ 2 وθ1 )
"!
رات ا
"ومJ& H %
r2 وr1 راتJ!
\ ا21";.و
r1 =
(
−θˆ1 1 − θˆ2
,
)
r =
−θˆ2
1 + θˆ12 + θˆ22 1 + θˆ12 + θˆ22
2
. θ 2 − θ1 < 1, θ 2 + θ1 < 1, θ 2 < 1 J%C 7;
ل ا1%
اA{ وθˆ2 وθˆ1 & H3
H% و
ARMA(1,1) ذج7 -5
zt − µ = φ1 ( zt −1 − µ ) + at − θ1at −1 , at ∼ N ( 0,σ 2 )
تB "
;[م اφ1 وθ1 )
"!
رات ا
"ومJ& د29
ρ1 =
(1 − φ1θ1 )(φ1 − θ1 ) , ρ = (1 − φ1θ1 )(φ1 − θ1 ) φ
1 + θ1 − 2φ1θ1 1 + θ1 − 2φ1θ1
2 2 2 1
φ1 وθ1 )
"!
رات ا
"ومJ& H %
r2 وr1 راتJ!
\ ا21";.و
r1 =
(1 − φˆθˆ )(φˆ − θˆ ) ,
1 1 1 1
r2 =
(1 − φˆθˆ )(φˆ − θˆ ) φˆ
1 1 1 1
1 + θˆ − 2φˆ1θˆ1 1 + θˆ − 2φˆ1θˆ1
2 2 1
1 1
r1 رJ!
4N"!
ا4
ا
!"دr2 رJ!
4N"!
ا4
ا
!"د4! J.و
r2
φˆ1 =
r1
4N"!
ا4
ا
!"د7N φˆ1 ض1"
θ1 )"!
ر ا
"ومJ& د29 . φ1 )"!
ر ا
"ومJ& 1وه
r1 رJ!
r2 ˆ r2 ˆ
1 − r θ1 r − θ1
r1 = 1 1
r
1 + θˆ12 − 2 2 θˆ1
r1
89
. θˆ1 < 1 J%C 7;
ا4!J
اA{
وθˆ1 رJ!
4 C'
ا4".;
ا4
ا
!"دH%
و
4
;
رات ا
"وم
!"
) ا
'!ذج اJ& L أو:#ر
ARIMA(1,1,1), ARIMA(2,1,0), ARIMA(0,1,2), ARIMA(1,2,0),
ARIMA(0,2,1), ARIMA(0,2,0).
.4Bرات أآ] دJ& د29 4
) أوJ ;[م آC رات ا
"ومJ& : !K,
: Conditional Least Square Method i
ا7 ت ا$
اi :7]
H3K
اW;3C 7;
واARMA(p,q)
'!ذج
φ p ( B )( zt − µ ) = θ q ( B ) at , at ∼ N ( 0,σ 2 )
O"!L VJC θ q ( B ) = 0 4
ور ا
!"دAL و4;آK& ورAL O'. L12X θ q ( B ) وφ p ( B ) Q5
:7
;
آat ء6
.
ذج ا1!'
ا4.;دة آF. .( بJ
9ة ) ط ا51
ة ا8رج دا
φp (B)
at = ( zt − µ )
θq ( B)
µ وθ = {θ1 ,θ 2 ,… ,θ q } وφ = {φ1 ,φ2 ,… ,φ p } )
"!
ا7N 4
إ;(رة آا3!2 !2ف ا6
ا
W;32 و
a ( φ, θ, µ ) =
(1 − φ B − φ B
1 2
2
−⋯ − φpB p )
( zt − µ )
t
(1 − θ B − θ B
1 2
2
−⋯ −θ pB p )
Y C z = {z1 , z2 ,… , zn } ة6"& هاتK!
و4=K
"ت ا
ا.!
ا4J2= !;"C
4
ا
ا
n
min Sc ( φ, θ, µ ) =
φ ,θ , µ
∑ a ( φ, θ, µ z )
t = p +1
2
t
.راتJ!
4( '
. 4
;
ا4 C'
اNormal Equations 4"(6
ت اX ا
!"دH5و
90
∂ ∂ n 2
∂φ
Sc ( φ, θ, µ )
φ =φˆ
= ∑ at ( φ, θ, µ z ) φ=φˆ = 0
∂φ t = p +1
θ=θ
ˆ ˆ θ= θ
µ = µˆ µ = µˆ
∂ ∂ n 2
∂θ
Sc ( φ, θ, µ )
φ =φˆ
= ∑ at ( φ, θ, µ z ) φ=φˆ = 0
∂θ t = p +1
θ =θˆ θ =θˆ
µ = µˆ µ = µˆ
∂ ∂ n 2
∂µ
Sc ( φ, θ, µ )
φ= φˆ
=
∂µ
∑ at ( φ, θ, µ z ) φ=φˆ = 0
t = p +1
θ =θ
ˆ θ = θˆ
µ = µˆ µ = µˆ
42 أي & وa p = a p −1 = ⋯ = a p +1−q = 0 )J
;ط ان اK ' ' ه4= ! C راتJ!
@ اAه
.( t = p + 1 4!J
(أ & ا2 4J.
ت اX ا
!"د7N V! ;
أن ا5X ) .O"B1;
& σ 2 2(;
ر اJ2
σˆ 2 =
(
Sc φˆ , θˆ , µ )
n − ( p + q + 1)
: اذجv i ا7 ت ا$ ات ا
AR(1) ذج7 -1
zt − µ = φ1 ( zt −1 − µ ) + at , at ∼ N ( 0,σ 2 )
z رهJ!. µ ف ;(ل1 تBJ;9 اw (;
zt − z = φ1 ( zt −1 − z ) + at , at ∼ N ( 0,σ 2 )
ء6 اW;3 z = {z1 , z2 ,… , zn } ة6"& هاتK!
at (φ1 ) = ( zt − z ) − φ1 ( zt −1 − z ) , t = 2,3,⋯ , n
هاتK!
اH آV!
واN6
اV.Cو
at2 (φ1 ) = ( zt − z ) − φ1 ( zt −1 − z ) , t = 2,3,⋯ , n
2
n n
Sc (φ1 ) = ∑ at2 (φ1 ) = ∑ ( zt − z ) − φ1 ( zt −1 − z )
2
t =2 t =2
b
42 & و4 ;'
ن ا13C وφ1 )"!
4( '
. 4J.
ا4
; ا
!"دK
، wJN φ1 )"!
4
@ داAوه
أيφ1 = φˆ1 &'
91
n n
Sc (φ1 ) = ∑ at2 (φ1 ) = ∑ ( zt − z ) − φ1 ( zt −1 − z )
2
t =2 t =2
∂ ∂ n
Sc (φ1 ) = ∑ ( zt − z ) − φ1 ( zt −1 − z )
2
∂φ1 ∂φ1 t =2
n
= ∑ −2 ( zt −1 − z ) ( zt − z ) − φ1 ( zt −1 − z )
t =2
∂ n
Sc (φ1 ) ˆ = ∑ −2 ( zt −1 − z ) ( zt − z ) − φˆ1 ( zt −1 − z ) = 0
∂φ1 φ1 =φ1
t =2
n
∴ ∑ ( zt −1 − z ) ( zt − z ) − φˆ1 ( zt −1 − z ) = 0
t =2
n n
∑(z − z )( zt − z ) − φˆ1 ∑ ( zt −1 − z ) = 0
2
t −1
t =2 t =2
أي
n
∑(z t −1 − z )( zt − z )
φˆ1 = t =2
n
∑(z −z)
2
t −1
t =2
. φ1 )"!
4=K
"ت ا
ا.!
ر اJ& 1وه
. φ1 )"!
ر ا
"ومJ&ر وJ!
ا اA ه. رنB : 2!C
MA(1) ذج7 -2
zt − µ = at − θ1at −1 , at ∼ N ( 0, σ 2 )
w1;!
4
"!
ا4 ;!
اH!"
و z رهJ!. µ ف
;(ل1 تBJ;9 اw (;
ذج1!'
(^ اN xt = zt − z
xt = at − θ1at −1 , at ∼ N ( 0, σ 2 )
H3K
اة ا4
ا
!"د4.;3.و
at = xt − θ1at −1
ء6 اW;3 = a0 = 0 VP1. وx1 , x2 ,… , xn ة6"& هاتK!
و
a1 = x1
a2 = x2 − θ1a1
a3 = x3 − θ1a2
⋮
an = xn − θ1an−1
92
7
;
.و
n
Sc (θ1 ) = ∑ at2
t =1
Q%(
ق ا6. Sc (θ1 ) Y C 7;
واθ1 4!B د2 إ3!2 وθ1 )"!
ا7N 46 z 4J.
ا4
ا
ا
7N o[;C 7;
واC1
-وسL 4J2= ( أو إ;[ام-1,1) ا
! ل7N 73(K
اQ%(
اH]& 42ا
"د
&] أيθ * 4
أو4!B ل15 θ1 )"!
46 4
ا. at = at (θ1 ) W2JC
dat (θ * )
at (θ1 ) ≈ at (θ *
) + (θ −θ *
) dθ1
1
dat (θ * )
4( '
. at = xt − θ1at −1 4
ا
!"د7N= قJ;F. f
وذ2ار3C O. 5 3!2 4J;K!
ا
dθ1
H %'
θ1 )"!
dat (θ1 ) θ1dat −1 (θ1 )
= + at −1 (θ1 )
d θ1 dθ1
da0 (θ1 )
4
ا
!"د. = 0 4
أو4!J. و
dθ1
dat (θ * )
at (θ1 ) ≈ at (θ * ) + (θ1 − θ * )
dθ1
"ت.!
ع ا1! & Y C ن3&X. 7
;
. وθ1 )"!
ا7N 46
n
Sc (θ1 ) = ∑ at2
t =1
رJ!
. θ * ;(الF. 4!"
@ اAر ه3
وθ1 )"!
H`N وأ2L رJ& 7 H %'
%C
"ت.!
ع ا1! & 7N oJ'
ا أو اL Y+
C 2رJ& . قb
(^ ا2 ;5 !;
و2
ا
ربJC H %
73
θ * 4
أو4!J
د ا29 ا
"وم4J2= إ;[ام3!& .اL Y+
.f
A
W5
;ج إ%C H. 2و2 );CX 4J.
ا4J26
=(" ا.V2
7;
ا46;[!
ك او ا
'!ذج ا%;!
اw1;!
!ذج ا4
5 7N ذج1!'
)
"!
ا2JC أن5 2
76 z H3K. ك%;!
اw1;!
ى &"
) ا1%C O
اJ"C H3KC ك%;& w1;& ي1%C
."!L O6 . أ1 وهMA(1) ذج1!'
ا4
5 7N
آ! هO%
42;ج إ
=ق د%C اAO
و
93
4&[; !
ى اXق ا6
"\ ا. آA
3
و4J.
; اJ26
ا اAر
هJ& 7N 7b;3
ف1
:H]& ذج1!'
&"
) ا2JC 7N
Maximum Likelihood Method !I"
ا4%L ار4J2= -1
Unconditional Least Squares Method 4=K
اz
"ت ا.!
ا4J2= -2
Nonlinear Estimation Methods 46[
اz2J;
=ق ا-3
94
>:وإ?ر ا ذج : Model Checking and Diagnostics
".ا
;"ف
!1ذج &( 78و )
"& 2JCهAا ا
'!1ذج
ي \".ا
; [Kت ا
(1ا7B
أو أ6ء ا
;) (6ا
'
(4 e2"C Iى &ى & 4J.6ا
'!1ذج
!; 4ا
!Kهة ،و;b2ض
أن ا
(1ا 7BهJ& 7رات
!; 4ا
` 4ا
(`ء atوا
;;b
7ض ا
1& Oزw1;!. "(= 4
b+ي و . σ 2 2(Cا
(1ا4B "
. 6"C 7B
et = zt − zˆt = aˆt , t = 1, 2,..., n
أي ان ا
(1ا 7Bه 7ا
)Jا
!Kهة
oBا
)Jا
!.4J(6
1J2م ا
; o[Kوا(;9رات o%Nا
(1ا 7Bوا
'& 7N Iى Pb
OJJ%Cت ا
'!1ذج
وا
; 7ه:7
b+ w1;& -1ي
-2ا
"1Kا48
-3م ا
;اw.
σ2 و2(C b+ي w1;!. و&;.6 )& ;HJ =("7 1CزV2 &1ز4 -4
أي ) ( at ∼ IIDN ( 0,σ 2
AOا
'
FNي o[KCوه & 41! & 1ا(;9رات 7ا
(1ا'
7Bي !Nإذا آ
J%C S
ه @Aا
Kوط و 7Nه @Aا
(;"
4
%ا
'!1ذج ا
! X1(J& (6أ& إذا HKNا 5ه @Aا(;9رات
' W Nإدة ا
' Iوإ;Bاح
!1ذج ,
أو :rإ?ر ا :\0
H 0 : E ( at ) = 0
H 1 : E ( at ) ≠ 0
e
= uوا
;1C O
7ز'"N 7B 7"(= V2 وه 1إ;(ر 2A.و
;[م 4Nا48 59
) se ( e
& ;1ى &"' (;"
α = 0.05 421ان E ( at ) = 0إذا آ
) u < 1.96 SهAا 7إ;(ر ان
) 5ا
"' 4اآ( & 30و5ة وهAا دا ;!
J%;& !8ت ا
&' 4ا
;
7ر( O
95
: إ?ر ا ا:7]
b
ل ا15 وw1;!
ل ا15 Runs test إ;(ر ا
ي46ا1. 7Bا1(
ا48ا1K (;[
7N W
6
اOر2 48ا1K"
;(رات9 آ] & اL12 ) 4!"&
;(رات ا9 ا5 ا1وه
.(;(ر9ا اAO. ' ه7b;3
3
وQ%. 241 رJ!
ا
:ل, 0j\ أو ا$ إ?ر ا
ا:-]
Autocorrelation 7CاA
اw. إ;(ر ا
;ا46ا1. 7Bا1(
ل اJ; أو إw.اC (;[2
4
داV& O;
رJ& و7Bا1(
SACF 4'"
ا4CاA
ت ا6. ب ور) ا
;ا%. f
وذtest
. ا
(`ء4 `
ا4 ;!
7CاA
اw.ا
;ا
;(ر9ا
H 0 : ρ1 = 0
H1 : ρ1 ≠ 0
r1
(;"
α = 0.05 421'"& ى1; & '"N 7B 7"(= V2ز1C O
u = 48 59 اQ5
se ( r1 )
. u < 1.96 S إذا آρ1 = 0 ان
:I ا اi إ?ر:$را
:H]& "ة =ق. f
=(" وذ4ز1& 7Bا1(
اS & إذا آ7N (;[
7!"&
;(ر ا9 و ;[م اGoodness of Fit Test .6;
ا5 إ;(ر-1
. Kolmogorov-Smirnov Test ف1 ! -روف1L1!
1آ
. Normal Probability Plot 7"(6
;!ل ا59 اw6[& -2
. Q-Q Plot "ت.
ا-"ت.
اw6[& -3
:N0 ذج ا7 ?رj ? ىr ا اv$
;[مC وLBQ ;[C وLjung-Box Q statistc ~آ1.-'
ـ1 آ48 5( إ1
:4Pb
;(ر ا9
H 0 : ρ1 = ρ 2 = ⋯ = ρ K = 0
:4B "
. 6"Cو
K
rk2
Q = n (n + 2) ∑ ∼ χ 2 ( K − m)
k =1 n − k
.ذج1!'
ا7N رةJ!
د ا
!"
) اm Q5
96
6"C وAIC ;[C وAutomatic Information Criteria 7CاA
&ت ا1"!
( &"ر ا2
:4B "
.
AIC ( m ) = n ln σ a2 + 2m
min AIC ( m ) 76"2 يA
ذج ا1!'
ذج و [;ر ا1!'
ا7N رةJ!
د ا
!"
) اm Q5
m
: Examples and Case Studies 0ت دراrK و-
أ
هةK& 4'& ز4 ;!
4
;
ا
( ت ا-1
z(t)
60.1815 59.5257 58.9275 56.4828 56.1346 57.2318 60.7196 59.9315
61.0640 61.4230 63.1547 63.9622 63.5049 64.6886 62.8556 61.0344
58.0059 58.7108 57.9813 59.1721 62.4654 60.5820 59.3191 60.6643
61.2223 61.4761 61.1856 60.9225 59.3054 58.3755 59.5353 60.5777
61.9753 62.1789 61.8108 58.1483 58.4174 60.1325 59.6004 59.9086
60.4833 61.7008 59.1609 59.4554 59.0903 58.0151 59.1455 62.2658
63.4411 60.5918 65.1325 61.7122 58.8802 59.5333 60.9492 61.9013
59.3478 59.4444 62.6899 61.6708 63.7261 55.7339 58.1690 54.5045
56.7241 57.3334 57.9363 58.5870 61.8370 58.9585 56.7437 55.8451
58.1281 62.1017 59.9443 60.2990 61.6337 61.1520 63.8189 59.3572
61.7840 57.3292 54.7163 58.2273 58.7564 59.0087 59.3402 61.8956
60.9021 63.1070 60.0538 63.6776 60.8942 60.5289 59.9246 59.7252
60.7001 58.1895 54.5550 54.6083 56.5413 59.1567 57.9624 58.4651
61.9462 61.9205 63.3933 62.3827 61.4310 60.3373 57.8803 61.2797
61.9448 56.2599 59.9569 57.8763 59.2086 55.4219 54.2185 58.0143
60.9805 62.1362 60.0855 60.3843 60.8605 62.3728 57.0642 56.6085
57.5151 58.4221 60.6919 63.5907 61.4451 60.1458 57.3940 56.8697
59.2145 60.8962 61.1852 58.1711 53.8560 57.5307 59.3236 57.2961
58.5278 60.3030 60.6201 59.9346 59.4119 61.5614 61.1107 59.6266
60.3550 60.7021 60.7227 58.0423 59.3488 60.0377 58.7336 58.1105
59.4242 58.5790 58.6501 55.4010 59.3839 60.8256 62.1957 61.9152
60.3319 57.1459 59.0970 59.0997 59.8597 59.0780 56.9972 59.0778
97
61.5555 60.9815 60.3563 59.5097 58.3583 63.1777 61.8685 58.2759
59.7755 60.2052 60.2513 59.2927 56.1494 56.0309 56.6666 59.5015
59.4755 60.9013 61.2179 61.1168 61.7218 59.2298 60.7356 63.4124
MINITAB 48 59 ا4&%
;[ام اF. Time Plot 7'& زw6[& 7N 4 ;!
) اXاو
:7
;
آ
MTB > TSPlot 'z(t)';
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect;
SUBC> Title "An obseved Time Series".
A n o b s e v e d T im e S e r ie s
65
60
z(t)
55
In d e x 50 100 150 200
&;[ام اF. 7'"
ا7CاA
اw. و ) ا
;اW % U
MTB > %ACF 'z(t)';
SUBC> MAXLAG 20;
SUBC> TITLE"SACF of observed Time Series".
Executing from file: H:\MTBWIN\MACROS\ACF.MAC
98
S A C F o f o b s e rv e d T im e S e rie s
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 10 15 20
L a g C o rr T LBQ L a g C o rr T LBQ L a g C o rr T LBQ
1 0 .5 1 7 .1 9 5 2 .4 8 8 -0 .1 4 -1 .5 0 8 1 .7 6 15 0 .0 7 0 .6 8 9 1 .2 3
2 0 .2 0 2 .3 2 6 0 .7 8 9 -0 .1 4 -1 .5 2 8 6 .1 4 16 0 .1 3 1 .3 3 9 4 .8 6
3 -0 .0 0 -0 .0 1 6 0 .7 8 10 -0 .0 9 -0 .9 0 8 7 .7 3 17 0 .1 7 1 .7 5 1 0 1 .3 3
4 -0 .0 5 -0 .5 9 6 1 .3 4 11 -0 .0 7 -0 .7 1 8 8 .7 3 18 0 .2 0 2 .0 6 1 1 0 .6 3
5 -0 .0 8 -0 .9 5 6 2 .8 2 12 -0 .0 8 -0 .7 9 8 9 .9 7 19 0 .1 2 1 .2 1 1 1 3 .9 8
6 -0 .1 8 -2 .0 5 6 9 .9 2 13 -0 .0 2 -0 .2 1 9 0 .0 5 20 0 .0 6 0 .6 1 1 1 4 .8 6
7 -0 .1 9 -2 .0 9 7 7 .5 8 14 0 .0 3 0 .3 2 9 0 .2 7
&;[ام اF. 7'"
ا78
ا7CاA
اw. و ) ا
;اW % ]
U
MTB > %PACF 'z(t)';
SUBC> MAXLAG 20;
SUBC> TITLE"SPACF of obseved Time Series".
Executing from file: H:\MTBWIN\MACROS\PACF.MAC
S P A C F o f o b se ve d T im e S e rie s
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 10 15 20
L a g P AC T L a g P AC T L a g P AC T
1 0 .5 1 7 .1 9 8 -0 .0 1 -0 .1 2 15 -0 .0 2 -0 .2 3
2 -0 .0 8 -1 .0 7 9 -0 .1 2 -1 .6 3 16 0 .0 9 1 .2 8
3 -0 .1 0 -1 .3 8 10 0 .0 1 0 .1 6 17 0 .0 8 1 .2 0
4 0 .0 0 0 .0 4 11 -0 .0 4 -0 .6 0 18 0 .0 6 0 .8 6
5 -0 .0 5 -0 .7 3 12 -0 .0 9 -1 .3 4 19 -0 .0 3 -0 .4 5
6 -0 .1 6 -2 .3 4 13 0 .0 3 0 .3 9 20 0 .0 4 0 .5 2
7 -0 .0 4 -0 .5 0 14 0 .0 2 0 .3 2
ذج1! V(;C 4 ;!
ان ا5 7'"
ا78
ا7CاA
اw. و ا
;ا7CاA
اw.& أ !ط ا
;ا
&;[ام اF. هاتK!
ا7 ;حJ!
ذج ا1!'
( ا6 اAO
وAR(1)
MTB > Name c7 = 'RESI1'
MTB > ARIMA 1 0 0 'z(t)' 'RESI1';
99
SUBC> Constant;
SUBC> Forecast 5 c4 c5 c6;
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot.
ARIMA Model
ARIMA model for z(t)
Estimates at each iteration
Iteration SSE Parameters
0 839.667 0.100 53.870
1 746.819 0.250 44.876
2 695.840 0.400 35.883
3 685.086 0.502 29.769
4 685.054 0.507 29.458
5 685.054 0.507 29.443
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.5073 0.0611 8.30
Constant 29.4429 0.1309 224.98
Mean 59.7571 0.2656
Number of observations: 201
Residuals: SS = 685.020 (backforecasts excluded)
MS = 3.442 DF = 199
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 10.8(DF=11) 27.6(DF=23) 35.9(DF=35)
45.0(DF=47)
100
Forecasts from period 201
95 Percent Limits
Period Forecast Lower Upper
Actual
202 59.7079 56.0707 63.3451
203 59.7322 55.6537 63.8106
204 59.7445 55.5600 63.9290
205 59.7507 55.5394 63.9620
206 59.7539 55.5357 63.9721
:7
;
و
;';_ ا
1;ح هJ!
ذج ا1!'
ا-1
zt = 59.76 + 0.51( zt −1 − 59.76) + at , at ∼ WN ( 0,3.44 )
:7 هO
t 4!B ا
!"ري وONا%
رة وإJ!
ا
!"
) ا-2
( )
φˆ1 = 0.51, s.e. φˆ1 = 0.061, t = 8.3
µˆ = 59.76, s.e. ( µˆ ) = 0.66
( )
δˆ = 29.44, s.e. δˆ = 0.131, t = 224.98
σˆ 2 = 3.44, with d . f . = 199
:7Bا1(
اo%b
".را
7Bا1(
اw1;& إ;(ر-1
MTB > ZTest 0.0 1.855 'RESI1';
SUBC> Alternative 0;
SUBC> GHistogram;
SUBC> GDotplot;
SUBC> GBoxplot.
Z-Test
Test of mu = 0.000 vs mu not = 0.000
The assumed sigma = 1.85
Variable N Mean StDev SE Mean Z P
RESI1 201 -0.002 1.851 0.131 -0.01 0.99
b
وي ا2 w1;!
{ن ا. 42b
ا4Pb
\ اN X
7Bا1(
ا48ا1K إ;(ر-2
101
MTB > Runs 0 'RESI1'.
Runs Test
RESI1
K = 0.0000
The observed number of runs = 94
The expected number of runs = 101.0796
107 Observations above K 94 below
The test is significant at 0.3149
Cannot reject at alpha = 0.05
48ا1K 7Bا1(
{ن ا. 42b
ا4Pb
\ اN X
7Bا1(
7CاA
اw. ا
;ا-3
ACF of Residuals for z(t)
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 10 15 20 25 30 35 40 45 50
Lag
7Bا1(
78
ا7CاA
اw. ا
;ا-4
102
PACF of Residuals for z(t)
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 10 15 20 25 30 35 40 45 50
Lag
ا
(`ء4 `
ا4 ;& أ
!طV(;C 78
ا7CاA
اw. وا
;ا7CاA
اw. ان أ
!ط ا
;ا5
: 7Bا1(
ا4"(= إ;(ر-5
7Bا1(
اري3;
اV`!
) ا-ا
Histogram of the Residuals
(response is z(t))
30
Frequency
20
10
-5 0 5
Residual
:
اI'
انW 2 H. 7b32X اA وه.(2JC 7"(6
اV2ز1;
اH3 4
أ
* &;' و5
Normal Probability Plot 7"(6
;!ل ا5X ر) ا-ب
103
Normal Probability Plot for RESI1
Mean: -1.6E-03
99
StDev: 1.85070
95
90
80
70
Percent
60
50
40
30
20
10
5
-5.0 -2.5 0.0 2.5 5.0
Data
:م1J
و
;{آ4"(= 7Bا1(
^ & ا
) أن اPوا
7Bا1(
ا4"(6
K-S Test ;(رF. -ج
MTB > %NormPlot 'RESI1';
SUBC> Kstest;
SUBC> Title "Normal Test for Residuals".
Executing from file: H:\MTBWIN\MACROS\NormPlot.MAC
Normal Test for Residuals
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-5 0 5
RESI1
Average: -0.0016272 Kolmogorov-Smirnov Normality Test
StDev: 1.85070 D+: 0.045 D-: 0.060 D : 0.060
N: 201 Approximate P-Value: 0.074
:7
;
ا5
و
1;(ر ه9ا
H 0 : Residuals ∼ N ( 0,3.44 )
H1 : Residuals§N ( 0,3.44 )
6ف ا1 !-روف1L1!
1إ;(ر آ
D + = 0.045, D − = 0.06, D = 0.06
104
4Pb
\ اN
X '
أي اα = 0.05 & ( أآ7 وه0.074 7
;(ر هP-Value ا
ـ
.42b
ا
:ات4
4(J; & )B 5 :(';
ذج1!'
ا;[&' ا
Forecasts from period 201
95 Percent Limits
Period Forecast Lower Upper Actual
202 59.7079 56.0707 63.3451
203 59.7322 55.6537 63.8106
204 59.7445 55.5600 63.9290
205 59.7507 55.5394 63.9620
206 59.7539 55.5357 63.9721
7
;
& ا. O!
و
Plot C4*C8 C5*C8 C6*C8;
SUBC> Connect;
SUBC> Type 1;
SUBC> Color 1;
SUBC> Size 1;
SUBC> Title "Forecast of 5 future value with 95%
limits";
SUBC> Overlay.
Forecast of 5 future value with 95% limits
64
63
62
61
60
C4
59
58
57
56
55
1 2 3 4 5
C8
105
.:(';
;ات اNات و:(';
اV& 4 ;!
ا
ء ا & ا76"2 7
;
وا
) ا
Forecast of 5 future value with 95% limits
64
63
62
61
60
C9
59
58
57
56
55
180 190 200
C8
هةK& 4'& ز4 ;!
4
;
ا
( ت ا-2
z(t)
499.148 496.650 511.026 488.539 498.440 507.382 496.208 494.948
503.975 501.649 489.348 506.040 496.678 502.233 498.429 503.170
498.758 502.969 498.229 501.605 493.371 505.884 496.227 496.806
493.057 506.459 502.545 497.785 506.329 496.665 491.923 504.340
499.890 494.559 503.107 502.891 498.598 500.074 499.260 496.372
507.416 500.508 496.830 491.981 516.373 492.286 500.356 503.506
498.090 498.319 507.020 493.161 499.217 508.489 494.033 496.062
504.877 498.304 495.355 505.581 495.000 504.965 497.393 501.521
494.918 501.527 504.712 501.064 492.352 500.664 495.431 507.886
499.173 494.833 504.072 497.883 495.423 507.072 496.285 506.345
496.765 504.129 495.737 500.744 505.577 485.991 507.673 507.735
482.567 507.594 503.580 493.866 501.819 500.921 503.415 497.295
500.989 498.294 501.700 495.868 501.175 503.852 499.783 497.642
501.331 496.932 507.582 494.885 504.666 498.380 496.181 510.287
489.314 504.394 501.928 494.814 509.407 498.060 497.133 496.029
502.720 499.982 503.325 495.954 504.408 500.199 494.878 503.134
502.489 498.640 500.484 493.552 501.417 504.785 497.943 501.634
495.691 502.173 502.066 497.130 492.318 505.517 499.299 499.611
106
496.252 504.346 501.082 497.626 496.757 505.475 498.787 500.388
499.279 504.913 493.843 506.259 498.403 497.462 499.467 505.987
498.169 500.712 498.571 504.085 491.707 504.817 502.933 493.858
497.015 504.204 501.703 490.683 505.429 504.336 495.430 494.857
503.195 506.403 498.599 487.344 514.220 490.887 511.741 497.861
500.252 502.721 500.256 494.614 502.414 503.465 501.999 493.017
498.158 503.746 497.643 507.438 491.418 506.649 496.078 498.931
500.409 506.001 490.619 512.122 496.150 505.218 497.413 497.794
496.225 501.827 500.324 505.367 498.016 498.477 495.353 513.900
491.726 496.063 499.779 504.012 501.542 496.680 499.134 504.717
489.032 505.709 497.956 497.231 507.590 491.202 503.130 502.209
500.024 493.502 502.681 505.234 497.647 495.699 504.174 497.992
505.194 497.421 502.823 496.877 504.640 492.716 501.701 501.387
499.574 497.048
ﺍﻭﻻ :ﺴﻭﻑ ﻨﺭﺴﻡ ﻓﻘﻁ 50ﻤﺸﺎﻫﺩﺓ ﻤﻥ ﻫﺫﻩ ﺍﻝﻤﺘﺴﻠﺴﻠﺔ
510
)z(t
500
490
In d e x 10 20 30 40 50
W % : Uو ) ا
;ا w.ا
Aا 7Cا
"':7
107
Autocorrelation Function for z(t)
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 10 15 20
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 -0.53 -8.38 71.03 8 0.10 1.19 87.98 15 0.05 0.60 113.33
2 -0.05 -0.64 71.68 9 -0.11 -1.31 91.01 16 -0.02 -0.26 113.47
3 0.12 1.52 75.39 10 0.15 1.77 96.64 17 0.07 0.82 114.85
4 0.04 0.52 75.83 11 -0.18 -2.16 105.32 18 -0.10 -1.16 117.65
5 -0.13 -1.61 80.08 12 0.11 1.25 108.35 19 0.07 0.75 118.81
6 0.11 1.39 83.33 13 0.05 0.56 108.97 20 -0.06 -0.73 119.93
7 -0.09 -1.13 85.52 14 -0.12 -1.36 112.61
:7'"
ا78
ا7CاA
اw. و
) ا
;اW %
:]
U
Partial Autocorrelation Function for z(t)
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 10 15 20
Lag PAC T Lag PAC T Lag PAC T
1 -0.53 -8.38 8 0.04 0.65 15 0.03 0.49
2 -0.46 -7.28 9 -0.07 -1.09 16 -0.13 -2.09
3 -0.29 -4.52 10 0.12 1.94 17 0.06 1.01
4 -0.07 -1.08 11 -0.07 -1.15 18 -0.08 -1.32
5 -0.11 -1.71 12 -0.04 -0.71 19 0.03 0.44
6 0.02 0.30 13 0.08 1.19 20 -0.13 -2.07
7 -0.09 -1.36 14 -0.04 -0.59
ذج1!'
ا اA( ه6;. وMA(1) ذج1!
V(;C B 4 ;!
ان ا5
هةK!
& ا
!ط ا
MTB > Name c7 = 'RESI1'
MTB > ARIMA 0 0 1 'z(t)' 'RESI1';
SUBC> Constant;
SUBC> Forecast 5 c4 c5 c6;
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot.
ARIMA Model
108
ARIMA model for z(t)
Estimates at each iteration
Iteration SSE Parameters
0 6081.19 0.100 500.046
1 5265.34 0.250 500.004
2 4615.22 0.400 499.980
3 4109.70 0.550 499.967
4 3766.60 0.700 499.960
5 3727.32 0.841 499.959
6 3687.70 0.797 499.963
7 3687.08 0.790 499.962
8 3687.07 0.791 499.962
9 3687.07 0.790 499.962
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
MA 1 0.7905 0.0386 20.50
Constant 499.962 0.051 9708.40
Mean 499.962 0.051
Number of observations: 250
Residuals: SS = 3684.13 (backforecasts excluded)
MS = 14.86 DF = 248
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 26.7(DF=11) 35.9(DF=23) 63.1(DF=35)
82.8(DF=47)
Forecasts from period 250
95 Percent Limits
Period Forecast Lower Upper Actual
251 502.256 494.700 509.812
252 499.962 490.330 509.593
253 499.962 490.330 509.593
254 499.962 490.330 509.593
109
255 499.962 490.330 509.593
:7
;
و
;';_ ا
1;ح هJ!
ذج ا1!'
ا-1
zt = 499.962 + at − 0.7905at −1 , at ∼ WN ( 0,14.86 )
:7 هO
t 4!B ا
!"ري وONا%
رة وإJ!
ا
!"
) ا-2
( )
θˆ1 = 0.7905, s.e. θˆ1 = 0.0386, t = 20.50
( )
µˆ = δˆ = 499.962, s.e. δˆ = 0.051, t = 9708.40
σˆ 2 = 14.86, with d . f . = 248
:7Bا1(
اo%b
".را
7Bا1(
اw1;& إ;(ر-1
MTB > ZTest 0.0 3.847 'RESI1';
SUBC> Alternative 0.
Z-Test
Test of mu = 0.000 vs mu not = 0.000
The assumed sigma = 3.85
Variable N Mean StDev SE Mean Z P
RESI1 250 -0.007 3.847 0.243 -0.03 0.98
b
وي ا2 w1;!
{ن ا. 42b
ا4Pb
\ اN X
7Bا1(
ا48ا1K إ;(ر-2
MTB > Runs 0 'RESI1'.
Runs Test
RESI1
K = 0.0000
110
The observed number of runs = 134
The expected number of runs = 125.9920
126 Observations above K 124 below
The test is significant at 0.3103
Cannot reject at alpha = 0.05
48ا1K 7Bا1(
{ن ا. 42b
ا4Pb
\ اN
X
7Bا1(
7CاA
اw. ا
;ا-3
A C F o f R e s id u a ls fo r z (t)
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e a u t o c o r r e l a t io n s )
1 .0
0 .8
0 .6
Autocorrelation
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 10 15 20 25 30 35 40 45 50 55 60
Lag
7Bا1(
78
ا7CاA
اw. ا
;ا-4
P A C F o f R e s id u a ls f o r z (t)
( w ith 9 5 % c o n f id e n c e l im it s f o r t h e p a r t ia l a u to c o r r e l a t io n s )
1 .0
0 .8
Partial Autocorrelation
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 10 15 20 25 30 35 40 45 50 55 60
Lag
ا
(`ء4 `
ا4 ;& أ
!طV(;C 78
ا7CاA
اw. وا
;ا7CاA
اw. ان أ
!ط ا
;ا5
: 7Bا1(
ا4"(= إ;(ر-5
7Bا1(
اري3;
اV`!
) ا-ا
111
H istogram of the R esiduals
(resp on se is z(t))
30
Frequency
20
10
-10 0 10
R esidual
:
اI'
انW 2 H. 7b32X اA وه.(2JC 7"(6
اV2ز1;
اH3 4
أ
* &;' و5
Normal Probability Plot 7"(6
;!ل ا5X ر) ا-ب
Normal Probability Plot for RESI1
Mean: -6.9E-03
99 StDev: 3.84651
95
90
80
70
Percent
60
50
40
30
20
10
5
-10 -5 0 5 10
Data
:م1J
و
;{آ4"(= 7Bا1(
^ & ا
) أن اPوا
7Bا1(
ا4"(6
K-S Test ;(رF. -ج
MTB > %Qqplot 'RESI1';
SUBC> Conf 95;
SUBC> Ci.
Executing from file: H:\MTBWIN\MACROS\Qqplot.MAC
Distribution Function Analysis
Normal Dist. Parameter Estimates
112
Data : RESI1
Mean: -6.9E-03
StDev: 3.84651
MTB > %NormPlot 'RESI1';
SUBC> Kstest.
Executing from file: H:\MTBWIN\MACROS\NormPlot.MAC
Normal Probability Plot
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-10 0 10
RESI1
Average: -0.0069004 Kolmogorov-Smirnov Normality Test
StDev: 3.84651 D+: 0.034 D-: 0.051 D : 0.051
N: 250 Approximate P-Value: 0.105
6ف ا1 !-روف1L1!
1إ;(ر آ
D + = 0.034, D − = 0.051, D = 0.051
ان4Pb
\ اN
X '
أي اα = 0.05 & ( أآ7 وه0.105 7
;(ر هP-Value ا
ـ
."(= 4ز1& 7Bا1(
ا
:ات4
4(J; & )B 5 :(';
ذج1!'
ا;[&' ا
Forecasts from period 250
95 Percent Limits
Period Forecast Lower Upper Actual
251 502.256 494.700 509.812
252 499.962 490.330 509.593
253 499.962 490.330 509.593
254 499.962 490.330 509.593
255 499.962 490.330 509.593
:('C 95% ;اتN V& ات:(';
ا76"2 7
;
وا
) ا
113
Forecast of 5 future values with 95% limits
510
C4
500
490
1 2 3 4 5
C8
هةK& 4'& ز4 ;!
4
;
ا
( ت ا-3
z(t)
229.574 227.346 230.260 229.903 226.778 226.641 226.760 224.678
224.077 225.772 223.390 222.482 221.562 222.515 224.063 227.500
230.713 234.323 236.033 236.488 232.308 229.136 225.663 221.632
215.405 213.619 217.433 223.408 232.653 239.577 238.463 234.178
228.758 221.484 217.123 218.067 222.156 227.621 232.209 233.005
234.678 236.419 235.744 229.359 229.331 229.564 230.102 232.432
234.155 233.918 235.767 234.668 231.319 231.633 231.121 228.189
227.075 226.765 224.927 225.721 225.734 227.982 229.848 231.718
230.421 228.200 228.472 230.888 230.122 227.859 223.115 222.468
224.663 225.799 228.227 229.851 228.225 228.618 228.418 231.163
233.335 236.399 236.659 235.024 235.122 228.989 224.483 226.479
223.571 222.523 225.196 226.724 228.198 229.792 232.738 234.207
234.561 232.976 231.266 227.812 224.928 225.447 228.163 230.455
232.473 232.067 233.891 233.841 234.707 234.825 232.232 233.640
231.653 230.148 230.327 228.922 231.665 235.224 236.562 233.725
230.146 227.077 227.032 227.089 229.575 233.044 233.427 233.089
233.444 233.256 232.820 228.954 224.747 224.207 225.484 228.655
230.076 231.062 232.461 232.152 226.865 222.819 220.782 220.958
221.171 224.050 228.727 232.135 232.027 232.315 232.030 231.531
230.582 232.032 231.411 232.684 233.852 233.127 230.938 231.363
114
232.344 233.622 233.799 235.038 232.160 229.733 229.757 228.285
224.880 223.599 225.273 223.994 224.258 227.948 230.636 229.320
227.449 229.100 231.898 228.203 228.606 227.046 230.713 235.587
239.660 242.860 243.963 239.883 234.243 230.662 230.360 228.729
225.860 225.123 225.070 229.486 231.265 234.107 234.625 232.700
229.792 230.082 227.643 230.342 233.628 238.762 241.821 240.884
235.112 228.468 223.381 223.795 226.994 230.499 230.865 236.017
238.292 235.623 230.088 226.271 225.616 225.771 226.222 229.321
227.805 226.745 225.447 223.250 225.291 225.358 225.985 228.141
230.794 229.727 227.934 228.920 230.296 229.369 229.352 228.958
231.092 232.891 235.210 235.339 236.029 232.881 228.837 226.114
225.020 224.096
MINITAB 48 59 ا4&%
;[ام اF. Time Plot 7'& زw6[& 7N 4 ;!
) اXاو
(wJN هةK& 50) :7
;
آ
2 4 2
z(t)
2 3 2
2 2 2
In d e x 1 0 2 0 3 0 4 0 5 0
7'"
ا7CاA
اw. و
) ا
;اW %
U
115
A u t o c o r r e la t io n F u n c t io n f o r z ( t )
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 10 15 20
Lag C o rr T LBQ Lag C o rr T LBQ Lag C o rr T LBQ
1 0 .8 4 13 .2 7 1 7 8 .2 1 8 -0 .0 2 -0 .1 7 3 2 2 .0 5 1 5 -0 .2 1 -1 .6 5 3 7 6 .3 2
2 0 .5 1 5 .2 0 2 4 4 .2 9 9 0 .1 4 1 .1 8 3 2 7 .2 4 1 6 -0 .2 2 -1 .7 9 3 8 9 .9 3
3 0 .1 5 1 .3 5 2 4 9 .7 1 1 0 0 .2 3 1 .9 5 3 4 1 .4 8 1 7 -0 .1 8 -1 .4 2 3 9 8 .8 1
4 -0 .1 6 -1 .4 6 2 5 6 .2 1 1 1 0 .2 3 1 .8 8 3 5 5 .1 6 1 8 -0 .1 0 -0 .7 6 4 0 1 .3 7
5 -0 .3 3 -2 .9 8 2 8 3 .7 5 1 2 0 .1 5 1 .2 2 3 6 1 .1 7 1 9 0 .0 0 0 .0 2 4 0 1 .3 7
6 -0 .3 3 -2 .8 7 3 1 1 .2 3 1 3 0 .0 2 0 .1 5 3 6 1 .2 5 2 0 0 .0 8 0 .6 1 4 0 3 .0 5
7 -0 .2 0 -1 .7 3 3 2 1 .9 5 1 4 -0 .1 2 -0 .9 6 3 6 5 .0 4
7'"
ا78
ا7CاA
اw. و
) ا
;اW %
]
U
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r z ( t )
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 10 15 20
Lag PAC T Lag PAC T Lag PAC T
1 0 .8 4 1 3 .2 7 8 0 .0 7 1 .0 6 1 5 0 .0 4 0 .6 7
2 -0 .6 6 - 1 0 .3 9 9 -0 .0 7 -1 .1 2 1 6 -0 .0 6 -0 .9 3
3 -0 .0 7 -1 .1 3 1 0 0 .0 5 0 .7 3 1 7 0 .0 4 0 .6 0
4 -0 .0 6 -1 .0 0 1 1 -0 .0 7 -1 .1 8 1 8 -0 .0 7 -1 .1 4
5 0 .1 1 1 .7 8 1 2 0 .0 6 0 .9 0 1 9 0 .0 9 1 .4 4
6 0 .1 4 2 .1 8 1 3 -0 .1 5 -2 .4 1 2 0 -0 .0 9 -1 .3 9
7 -0 .0 1 -0 .2 2 1 4 0 .0 1 0 .1 8
ذج1! V(;C 4 ;!
ان ا5 7'"
ا78
ا7CاA
اw. و ا
;ا7CاA
اw.& أ !ط ا
;ا
هاتK!
ا7 ;حJ!
ذج ا1!'
( ا6 اAO
وAR(2)
MTB > Name c7 = 'RESI1'
MTB > ARIMA 2 0 0 'z(t)' 'RESI1';
SUBC> Constant;
SUBC> Forecast 10 c4 c5 c6;
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot.
116
ARIMA Model
ARIMA model for z(t)
Estimates at each iteration
Iteration SSE Parameters
0 4257.23 0.100 0.100 183.784
1 3528.31 0.250 0.012 169.535
2 2889.23 0.400 -0.076 155.360
3 2338.97 0.550 -0.165 141.201
4 1877.39 0.700 -0.253 127.051
5 1504.46 0.850 -0.342 112.913
6 1220.13 1.000 -0.430 98.789
7 1024.34 1.150 -0.519 84.690
8 916.97 1.300 -0.608 70.623
9 894.38 1.402 -0.668 61.154
10 894.31 1.408 -0.672 60.670
11 894.31 1.408 -0.672 60.646
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
AR 1 1.4079 0.0473 29.78
AR 2 -0.6720 0.0474 -14.19
Constant 60.6458 0.1203 504.11
Mean 229.638 0.456
Number of observations: 250
Residuals: SS = 893.567 (backforecasts excluded)
MS = 3.618 DF = 247
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36 48
Chi-Square 17.5(DF=10) 27.2(DF=22) 49.7(DF=34) 67.7(DF=46)
Forecasts from period 250
95 Percent Limits
Period Forecast Lower Upper Actual
251 224.939 221.211 228.668
252 226.747 220.308 233.186
117
253 228.725 220.642 236.808
254 230.296 221.546 239.045
255 231.177 222.311 240.044
256 231.363 222.494 240.233
257 231.033 222.070 239.996
258 230.442 221.327 239.558
259 229.833 220.600 239.067
260 229.372 220.090 238.655
:7
;
و
;';_ ا
1;ح هJ!
ذج ا1!'
ا-1
zt = 60.6458 + 1.4079 zt −1 − 0.672 zt − 2 + at , at ∼ WN ( 0,3.618)
:7 هO
t 4!B ا
!"ري وONا%
رة وإJ!
ا
!"
) ا-2
( )
φˆ1 = 1.4079, s.e. φˆ1 = 0.0473, t = 29.78
φˆ2 = −0.672, s.e. (φˆ ) = 0.0474,
2 t = −14.19
µˆ = 229.638, s.e. ( µˆ ) = 0.456
( )
δˆ = 60.6458, s.e. δˆ = 0.1203, t = 504.11
σˆ 2 = 3.618, with d . f . = 247
:I> ا ا/Y7 $را
7Bا1(
اw1;& إ;(ر-1
MTB > ZTest 0.0 3.618 'RESI1';
SUBC> Alternative 0.
Z-Test
Test of mu = 0.000 vs mu not = 0.000
The assumed sigma = 3.62
Variable N Mean StDev SE Mean Z P
RESI1 250 -0.005 1.894 0.229 -0.02 0.98
b
وي ا2 w1;!
{ن ا. 42b
ا4Pb
\ اN X
118
7Bا1(
ا48ا1K إ;(ر-2
MTB > Runs 0 'RESI1'.
Runs Test
RESI1
K = 0.0000
The observed number of runs = 125
The expected number of runs = 125.8720
129 Observations above K 121 below
The test is significant at 0.9119
Cannot reject at alpha = 0.05
48ا1K 7Bا1(
{ن ا. 42b
ا4Pb
\ اN X
7Bا1(
7CاA
اw. ا
;ا-3
A C F o f R e s id u a ls f o r z (t)
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e a u t o c o r r e l a t io n s )
1 .0
0 .8
0 .6
Autocorrelation
0 .4
0 .2
0 .0
- 0 .2
- 0 .4
- 0 .6
- 0 .8
- 1 .0
5 10 15 20 25 30 35 40 45 50 55 60
Lag
7Bا1(
78
ا7CاA
اw. ا
;ا-4
119
P A C F o f R e s id u a ls f o r z ( t)
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e p a r t ia l a u t o c o r r e l a t io n s )
1 .0
0 .8
Partial Autocorrelation
0 .6
0 .4
0 .2
0 .0
- 0 .2
- 0 .4
- 0 .6
- 0 .8
- 1 .0
5 10 15 20 25 30 35 40 45 50 55 60
Lag
ا
(`ء4 `
ا4 ;& أ
!طV(;C 78
ا7CاA
اw. وا
;ا7CاA
اw. ان أ
!ط ا
;ا5
: 7Bا1(
ا4"(= إ;(ر-5
7Bا1(
اري3;
اV`!
) ا-ا
H istogram of th e R esiduals
(resp on s e is z(t))
30
20
Frequency
10
-5 0 5
R es idual
:
اI'
انW 2 H. 7b32X اA وه.(2JC 7"(6
اV2ز1;
اH3 4
أ
* &;' و5
Normal Probability Plot 7"(6
;!ل ا5X ر) ا-ب
120
N orm al P rob ab ility P lot for R E S I1
M e a n: -4 .6 E -03
99 S tD e v : 1 .8 94 3 6
95
90
80
70
Percent
60
50
40
30
20
10
5
-6 -4 -2 0 2 4
D ata
:م1J
و
;{آ4"(= 7Bا1(
^ & ا
) أن اPوا
7Bا1(
ا4"(6
K-S Test ;(رF. -ج
Normal Probability Plot
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-5 0 5
RESI1
Average: -0.0046305 Kolmogorov-Smirnov Normality Test
StDev: 1.89436 D+: 0.020 D-: 0.029 D : 0.029
N: 250 Approximate P-Value > 0.15
& ( أآ7 وه0.15 7
;(ر هP-Value ا
ـ6ف ا1 !-روف1L1!
1إ;(ر آ
.7Bا1(
ا4"(= 4PN \N X ' أي اα = 0.05
:ات4
4(J; & )B 10 :(';
ذج1!'
ا;[&' ا
Forecasts from period 250
95 Percent Limits
Period Forecast Lower Upper Actual
251 224.939 221.211 228.668
252 226.747 220.308 233.186
121
253 228.725 220.642 236.808
254 230.296 221.546 239.045
255 231.177 222.311 240.044
256 231.363 222.494 240.233
257 231.033 222.070 239.996
258 230.442 221.327 239.558
259 229.833 220.600 239.067
260 229.372 220.090 238.655
:('C ;اتN 95% ات و:(';
ا76"2 7
;
وا
) ا
Forecast of 10 future values with 95% limits
240
C4
230
220
0 1 2 3 4 5 6 7 8 9 10
C8
Forecast of 10 future values with 95% limits
245
235
C9
225
215
0 100 200
C8
122
Forecast of 10 future values with 95% limits
240
C9
230
220
0 10 20 30 40 50 60
C8
V& اة4!B ![
7 ]
اH3K
ات وا:(';
اV& O&3. 4'&
ا4 ;!
( ا2 اولH3K
ا
.:(';
ا4
داH3 ^P1;
ات:(';
ا
: 0 دراK
ا6 أB & )إV' & إ
;جw 7N 4("!
ت ا12b;
"د ا7&1
اw1;!
ا4
;
ا4 ;!
ا
(6 .
Defects
1.20 1.50 1.54 2.70 1.95 2.40 3.44 2.83 1.76 2.00
2.09 1.89 1.80 1.25 1.58 2.25 2.50 2.05 1.46 1.54
1.42 1.57 1.40 1.51 1.08 1.27 1.18 1.39 1.42 2.08
1.85 1.82 2.07 2.32 1.23 2.91 1.77 1.61 1.25 1.15
1.37 1.79 1.68 1.78 1.84
4 ;!
7'&
اw6[!
ا
123
3 .5
3 .0
2 .5
Defects
2 .0
1 .5
1 .0
In d e x 10 20 30 40
78
ا7CاA
وا7CاA
اw.ا
;ا
Autocorrelation Function for Defects
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag Corr T LBQ Lag Corr T LBQ
1 0.43 2.88 8.84 8 -0.11 -0.57 17.25
2 0.26 1.49 12.18 9 -0.05 -0.27 17.41
3 0.14 0.77 13.18 10 -0.01 -0.04 17.41
4 0.08 0.43 13.50 11 -0.04 -0.19 17.50
5 -0.09 -0.46 13.89
6 -0.07 -0.39 14.18
7 -0.21 -1.10 16.57
Partial Autocorrelation Function for Defects
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag PAC T Lag PAC T
1 0.43 2.88 8 0.07 0.44
2 0.09 0.63 9 0.05 0.35
3 -0.00 -0.01 10 0.01 0.09
4 0.00 0.00 11 -0.03 -0.23
5 -0.16 -1.07
6 0.00 0.02
7 -0.18 -1.19
124
:4B "
. 6"2 يA
واAIC 7CاA
&ت ا1"!
ف
;[م &"ر ا1 W'!
ذج ا1!'
;ر ا9
AIC ( m ) = n ln σ a2 + 2m
min AIC ( m ) 76"2 يA
ذج ا1!'
ذج و [;ر ا1!'
ا7N رةJ!
د ا
!"
) اm Q5
m
:7
ا1;
ا7 ( ا
'!ذج6
ف1
MTB > ARIMA 1 0 0 'Defects' 'RESI1';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.4421 0.1365 3.24
Constant 0.99280 0.06999 14.19
Mean 1.7795 0.1254
Number of observations: 45
Residuals: SS = 9.47811 (backforecasts excluded)
MS = 0.22042 DF = 43
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 4.9(DF=11) 8.9(DF=23) 30.9(DF=35)
* (DF= *)
MTB > ARIMA 2 0 0 'Defects' 'RESI2';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
125
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.3999 0.1533 2.61
AR 2 0.0989 0.1531 0.65
Constant 0.89019 0.07047 12.63
Mean 1.7762 0.1406
Number of observations: 45
Residuals: SS = 9.38567 (backforecasts excluded)
MS = 0.22347 DF = 42
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 4.0(DF=10) 8.1(DF=22) 28.8(DF=34)
* (DF= *)
MTB > ARIMA 1 0 1 'Defects' 'RESI3';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.5983 0.2691 2.22
MA 1 0.1926 0.3294 0.58
Constant 0.71334 0.05693 12.53
Mean 1.7759 0.1417
Number of observations: 45
Residuals: SS = 9.39423 (backforecasts excluded)
MS = 0.22367 DF = 42
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
126
Lag 12 24 36
48
Chi-Square 4.1(DF=10) 8.3(DF=22) 29.1(DF=34)
* (DF= *)
MTB > ARIMA 0 0 1 'Defects' 'RESI4';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
Final Estimates of Parameters
Type Coef StDev T
MA 1 -0.3409 0.1431 -2.38
Constant 1.78480 0.09651 18.49
Mean 1.78480 0.09651
Number of observations: 45
Residuals: SS = 10.0362 (backforecasts excluded)
MS = 0.2334 DF = 43
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 8.0(DF=11) 13.2(DF=23) 35.7(DF=35)
* (DF= *)
MTB > ARIMA 0 0 2 'Defects' 'RESI5';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
127
Final Estimates of Parameters
Type Coef StDev T
MA 1 -0.3869 0.1516 -2.55
MA 2 -0.1816 0.1516 -1.20
Constant 1.7839 0.1118 15.96
Mean 1.7839 0.1118
Number of observations: 45
Residuals: SS = 9.61059 (backforecasts excluded)
MS = 0.22882 DF = 42
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 4.6(DF=10) 9.2(DF=22) 31.0(DF=34)
* (DF= *)
MTB > ARIMA 2 0 1 'Defects' 'RESI6';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.4134 1.5680 0.26
AR 2 0.0929 0.7113 0.13
MA 1 0.0136 1.5749 0.01
Constant 0.87675 0.07036 12.46
Mean 1.7761 0.1425
Number of observations: 45
Residuals: SS = 9.38561 (backforecasts excluded)
MS = 0.22892 DF = 41
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
128
Lag 12 24 36
48
Chi-Square 4.0(DF= 9) 8.1(DF=21) 28.8(DF=33)
* (DF= *)
MTB > ARIMA 1 0 2 'Defects' 'RESI7';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
* ERROR * Model cannot be estimated with these data
MTB > ARIMA 2 0 2 'Defects' 'RESI8';
SUBC> Constant;
ARIMA Model
ARIMA model for Defects
Final Estimates of Parameters
Type Coef StDev T
AR 1 1.6720 0.1165 14.35
AR 2 -0.7263 0.1251 -5.80
MA 1 1.3199 0.0184 71.63
MA 2 -0.3196 0.0731 -4.37
Constant 0.096224 0.003323 28.95
Mean 1.77238 0.06121
Number of observations: 45
Residuals: SS = 8.33225 (backforecasts excluded)
MS = 0.20831 DF = 40
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
129
Chi-Square 4.7(DF= 8) 8.9(DF=20) 29.7(DF=32)
* (DF= *)
:7
;
ول ا. f
ذo[
و
Model σˆ 2 m AIC
__________ ________ ___ _________
AR (1) 0.22042 3 −62.0499
AR ( 2 ) 0.22347 4 −59.4315
MA (1) 0.23340 3 −59.4751
MA ( 2 ) 0.22882 4 −58.3669
ARMA (1,1) 0.22367 4 −59.3913
ARMA ( 2,1) 0.22892 5 −56.3472
ARMA (1, 2 ) − − −
ARMA ( 2, 2 ) 0.20831 6 −58.5928
min AIC ( m ) = −62.0499
m
. AR(1) 1ذج ه1! H`Nأي ان أ
.ات:('C
1C و7Bا1(
اo%N 2!; آW
6
;ك2
130
Kدرا: 0
ا
!; 4ا
;
4ه 7د Hا
!("ت ا
' 2 !. 421ا
X2ت
Kآ& 4
Sales
3.49 5.74 5.51 3.99 3.45 4.77 4.14 4.60 3.80 5.43
3.96 2.54 4.05 6.16 3.78 5.07 5.42 3.91 4.30 3.88
2.89 4.61 4.08 4.05 3.28 2.65 1.22 3.98 3.45 3.57
2.52 1.58 4.00 5.14 3.84 4.40 3.08 5.43 4.80 2.75
5.77 4.99 4.31 6.46 6.11 4.79 5.65 5.52 6.12 6.06
3.20 5.05 6.23 6.12 4.99 4.89 4.78 5.67 6.08 5.80
5.13 7.07 8.02 6.36 5.75 5.70 5.61 5.63 5.71 5.16
7.20 6.87 7.56 6.57 6.08 4.72 6.09 6.64 7.49 6.64
7.26 7.22 6.69 7.49 9.01 7.27 5.62 7.59 7.53 6.43
6.42 8.22 7.67 7.53 7.23 8.50 8.27 8.75 7.50 7.86
&[ w6ز&'4 ;!
7
9
8
7
6
Sales
5
4
3
2
1
In d e x 10 20 30 40 50 60 70 80 90 100
دوال ا
;ا w.ا
Aا 7Cوا
Aا 7Cا
78ا
"'4
131
Autocorrelation Function for Sales
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 15 25
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 0.71 7.10 51.97 8 0.56 2.22 317.32 15 0.41 1.32 483.85 22 0.22 0.67 565.04
2 0.60 4.26 89.85 9 0.49 1.87 344.49 16 0.35 1.13 499.04 23 0.17 0.50 568.77
3 0.65 3.94 134.64 10 0.49 1.79 371.67 17 0.31 0.97 510.68 24 0.21 0.64 574.90
4 0.64 3.36 177.75 11 0.51 1.82 401.71 18 0.30 0.92 521.52 25 0.25 0.75 583.43
5 0.59 2.83 215.72 12 0.42 1.46 422.60 19 0.36 1.11 537.81
6 0.59 2.63 253.97 13 0.38 1.29 439.91 20 0.31 0.95 550.11
7 0.51 2.12 282.61 14 0.45 1.50 464.06 21 0.26 0.77 558.52
Partial Autocorrelation Function for Sales
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 15 25
Lag PAC T Lag PAC T Lag PAC T Lag PAC T
1 0.71 7.10 8 0.19 1.93 15 0.03 0.32 22 -0.17 -1.68
2 0.20 1.99 9 -0.17 -1.70 16 -0.10 -0.96 23 -0.10 -1.00
3 0.35 3.53 10 0.14 1.44 17 -0.08 -0.84 24 0.15 1.55
4 0.15 1.49 11 0.03 0.34 18 -0.01 -0.10 25 0.03 0.34
5 0.09 0.92 12 -0.15 -1.52 19 0.14 1.43
6 0.10 1.03 13 0.02 0.25 20 -0.04 -0.44
7 -0.13 -1.27 14 0.04 0.39 21 0.00 0.04
.w1;!
ا7N ارJ;ل م إ2 B !& |6. &[C لC 7'"
ا7CاA
اw. ا
;ا4
دا
O!
وwt = ∇zt 4 ;!
اول2b;
' ب ا
1
w(t)
-1
-2
-3
In d e x 10 20 30 40 50 60 70 80 90 100
132
ةJ; !
ا4 ;!
78
ا7CاA
وا7CاA
اw. دوال ا
;ا.نsة اJ; & 4 ;!
(و اC
Autocorrelation Function for w(t)
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 12 22
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 -0.30 -3.00 9.26 8 0.19 1.53 33.07 15 0.00 0.00 42.37 22 0.05 0.35 51.22
2 -0.31 -2.86 19.32 9 -0.07 -0.53 33.58 16 0.02 0.13 42.40 23 -0.18 -1.28 55.44
3 0.11 0.90 20.49 10 -0.07 -0.52 34.08 17 -0.07 -0.52 42.99 24 0.03 0.20 55.55
4 0.04 0.31 20.63 11 0.16 1.27 37.12 18 -0.15 -1.09 45.65
5 -0.04 -0.33 20.80 12 -0.07 -0.53 37.67 19 0.20 1.48 50.73
6 0.14 1.15 22.80 13 -0.14 -1.04 39.85 20 0.01 0.05 50.74
7 -0.24 -1.98 29.02 14 0.15 1.10 42.37 21 -0.04 -0.25 50.90
Partial Autocorrelation Function for w(t)
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 12 22
Lag PAC T Lag PAC T Lag PAC T Lag PAC T
1 -0.30 -3.00 8 0.12 1.17 15 0.06 0.59 22 0.08 0.80
2 -0.44 -4.41 9 -0.16 -1.59 16 0.07 0.70 23 -0.15 -1.53
3 -0.22 -2.23 10 -0.06 -0.60 17 -0.03 -0.32 24 -0.05 -0.55
4 -0.21 -2.05 11 0.09 0.93 18 -0.17 -1.65
5 -0.17 -1.72 12 -0.06 -0.59 19 -0.03 -0.31
6 0.06 0.59 13 -0.02 -0.17 20 -0.10 -0.95
7 -0.26 -2.55 14 -0.07 -0.71 21 0.09 0.85
:4B "
. 6"2 يA
واAIC 7CاA
&ت ا1"!
ف
;[م &"ر ا1 W'!
ذج ا1!'
;ر ا9
AIC ( m ) = n ln σ a2 + 2m
min AIC ( m ) 76"2 يA
ذج ا1!'
ذج و [;ر ا1!'
ا7N رةJ!
د ا
!"
) اm Q5
m
:7
ا1;
ا7 ( ا
'!ذج6
ف1
MTB > ARIMA 1 1 0 'Sales';
SUBC> NoConstant.
133
ARIMA Model
ARIMA model for Sales
Final Estimates of Parameters
Type Coef StDev T
AR 1 -0.3114 0.0959 -3.25
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 133.134 (backforecasts excluded)
MS = 1.359 DF = 98
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 31.9(DF=11) 51.2(DF=23) 62.8(DF=35)
81.0(DF=47)
MTB > ARIMA 2 1 0 'Sales';
SUBC> NoConstant.
ARIMA Model
ARIMA model for Sales
Final Estimates of Parameters
Type Coef StDev T
AR 1 -0.4532 0.0897 -5.05
AR 2 -0.4656 0.0901 -5.17
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 104.715 (backforecasts excluded)
MS = 1.080 DF = 97
134
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 21.8(DF=10) 40.9(DF=22) 49.4(DF=34)
59.9(DF=46)
MTB > ARIMA 0 1 1 'Sales';
SUBC> NoConstant.
ARIMA Model
ARIMA model for Sales
Final Estimates of Parameters
Type Coef StDev T
MA 1 0.7636 0.0648 11.78
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 101.411 (backforecasts excluded)
MS = 1.035 DF = 98
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 12.6(DF=11) 27.8(DF=23) 35.9(DF=35)
48.5(DF=47)
MTB > ARIMA 0 1 2 'Sales';
SUBC> NoConstant.
ARIMA Model
ARIMA model for Sales
135
Final Estimates of Parameters
Type Coef StDev T
MA 1 0.5756 0.0990 5.81
MA 2 0.2029 0.0998 2.03
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 99.2463 (backforecasts excluded)
MS = 1.0232 DF = 97
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 14.3(DF=10) 28.3(DF=22) 36.5(DF=34)
47.0(DF=46)
MTB > ARIMA 1 1 1 'Sales';
SUBC> NoConstant.
ARIMA Model
ARIMA model for Sales
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.1283 0.1334 0.96
MA 1 0.8027 0.0799 10.04
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 100.421 (backforecasts excluded)
MS = 1.035 DF = 97
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
136
Lag 12 24 36
48
Chi-Square 13.3(DF=10) 27.9(DF=22) 36.1(DF=34)
48.2(DF=46)
MTB > ARIMA 2 1 1 'Sales';
SUBC> NoConstant.
ARIMA Model
ARIMA model for Sales
* WARNING * Back forecasts not dying out rapidly
Final Estimates of Parameters
Type Coef StDev T
AR 1 -1.1389 0.0987 -11.53
AR 2 -0.1440 0.0983 -1.47
MA 1 -0.9889 0.0002 -3987.49
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 134.250 (backforecasts excluded)
MS = 1.398 DF = 96
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 35.1(DF= 9) 53.5(DF=21) 66.6(DF=33)
83.2(DF=45)
MTB > ARIMA 1 1 2 'Sales';
SUBC> NoConstant.
137
ARIMA Model
ARIMA model for Sales
Final Estimates of Parameters
Type Coef StDev T
AR 1 -0.3476 0.4077 -0.85
MA 1 0.2422 0.3771 0.64
MA 2 0.4506 0.2656 1.70
Differencing: 1 regular difference
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 97.2357 (backforecasts excluded)
MS = 1.0129 DF = 96
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 11.9(DF= 9) 25.0(DF=21) 32.1(DF=33)
41.8(DF=45)
MTB > ARIMA 2 1 2 'Sales';
SUBC> NoConstant.
ARIMA Model
ARIMA model for Sales
Final Estimates of Parameters
Type Coef StDev T
AR 1 -0.0691 0.3618 -0.19
AR 2 -0.2941 0.1450 -2.03
MA 1 0.5637 0.3737 1.51
MA 2 0.0840 0.3266 0.26
Differencing: 1 regular difference
138
Number of observations: Original series 100, after
differencing 99
Residuals: SS = 93.6368 (backforecasts excluded)
MS = 0.9857 DF = 95
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 11.0(DF= 8) 23.4(DF=20) 30.1(DF=32)
36.5(DF=44)
:7
;
ول ا. f
ذo[
و
Model σˆ 2 m AIC
__________ ________ ___ _________
ARI (1,1) 1.359 2 34.368
ARI ( 2,1) 1.080 3 13.619
IMA (1,1) 1.035 2 7.4057
IMA (1, 2 ) 1.023 3 8.2706
ARIMA (1,1,1) 1.035 3 9.4057
ARIMA ( 2,1,1) 1.398 4 41.169
ARIMA (1,1, 2 ) 1.013 4 9.2689
ARIMA ( 2,1, 2 ) 0.986 5 8.5741
min AIC ( m ) = 7.406
m
.ات:('C
1C و7Bا1(
اo%N 2!; آW
6
;ك2 . IMA(1,1) 1ذج ه1!
H`Nأي ان أ
139
ﺍﻟﻔﺼﻞ ﺍﻟﺴﺎﺩﺱ
0
ك ا/\ ا0 ا ا)
ا9 ار ا/7jذج ا7
Seasonal Autoregressive Integrated Moving Average Models
H]& "(
ا42 &; و4'&;ات زN ر3;C 4O.K;& ا
!ط76"C 4!1!
ا4'&
ا
!; ت ا
.4' اوO ا4U U اوO Hم او آ2 ا4"( H او آ4 ونK و4". ارH آw!'
ر ا3;2 ان
@ ا
!; تA هH]& (C 4
;
ل ا3ا
S e a s o n a l T im e S e r ie s
70
z(t)
60
50
In d e x 50 100 150
S e a s o n a l T im e S e r ie s
1000
900
800
z(t)
700
600
In d e x 50 100 150
46ا1. O;LA! و=ق4!1!
ا4'&
اص ا
!; ت ا1 ف ;"ض1 H b
ا اA ه7N
]!N SARIMA(p,d,q)(P,D,Q)s ك%;!
اw1;!
ا7&3;
ا7CاA
ار ا% 9 !ذج ا
H3K
اW;32 SARIMA(0,1,1)(1,1,0)12 ذج1!'
ا
140
(1 − Φ B ) (1 − B ) z = (1 − θ B ) a ,
1
s
t 1 t at ∼ WN ( 0, σ 2 )
(p,d,q)(P,D,Q)s 4L
ر. ك%;!
اw1;!
ا7&3;
ا7CاA
ار ا%
9ذج ا1!
نFN مH3K.و
H3K
اW;32 SARIMA(p,d,q)(P,D,Q)s
φ p ( B ) Φ P ( B s ) (1 − B ) (1 − B s ) zt = δ + θ q ( B ) ΘQ ( B s ) at , at ∼ WN ( 0,σ 2 )
d D
&ت7;
وا4!1!
اz ك%;!
اw1;!
وا7CاA
ار ا%
9 !ل اθ q ( B ) وφ p ( B ) Q5
و7!1!
ا7CاA
ار ا%
9 اH& Φ P ( B s ) = 1 + Φ1B s + Φ 2 B 2 s + ⋯ + Φ P B Ps وJ. '
اA ! ه2 و7!1!
ك ا%;!
اw1;!
اH& ΘQ ( B s ) = 1 + Θ1B s + Θ2 B 2 s + ⋯ + ΘQ B Qs
.Multiplicative Seasonal Models 7b`;
ا7!1!
ذج ا1!'
.
at ∼ WN ( 0, σ 2 ) !' أنP &1Ob& ن13 4&دJ
ا
'!ذج اV!L 7N :4I5 &
: 0 اذج اv Vا ا9\ ا$ا وا
ا9\ ا$دوال ا
ا
wt = (1 − B ) (1 − B s ) zt ةJ; !
ا4!1!
ا4 ;!
ف
;[م ا1 4
;
ت اBJ;9 ا7N
d D
SARMA(p,q)(P,Q)s ذج1!'
اV(;C 7;
وا
φ p ( B ) Φ P ( B s ) wt = δ + θ q ( B ) ΘQ ( B s ) at , at ∼ WN ( 0, σ 2 )
SARMA(0,1)(1,1)12 7b`;
ا7!1!
ذج ا1!'
7CاA
اw. ا
;ا4
; داK
ف1
wt = Φ wt −12 + at − θ at −1 − Θat −12 + θ Θat −13 , at ∼ WN ( 0,σ 2 )
VB1;
اA وأwt 7N 4N"!
ا4
ا
!"د7N= `ب.
γ 0 = Φ γ 12 + σ 2 + θ 2σ 2 − Θ ( Φ − Θ ) σ 2 + θ Θ ( −Φ θ + θ Θ ) σ 2
=Φ γ 12 + σ 2 (1 + θ 2 ) + Θ ( Φ − Θ ) (1 + θ 2 )
=Φ γ 12 + σ 2 (1 + θ 2 ) 1 + Θ ( Φ − Θ )
VB1;
اA وأwt −12 7N 4N"!
ا4
ا
!"د7N= `ب.و
γ 12 = Φ γ 0 − Θσ 2 + θ Θ ( −θ ) σ 2
=Φ γ 0 − Θσ 2 (1 + θ 2 )
;J.
; اB "
اH%.و
1 + Θ2 − 2ΦΘ
γ 0 = σ 2 (1 + θ 2 )
1 − Φ2
Φ (Θ − Φ )
2
γ 12 = σ (1 + θ ) Φ − Θ +
2 2
1 − Φ 2
141
`2أ
γ 1 = E ( wt wt −1 )
=Φ γ 11 − θσ 2 − ΘE ( at −12 wt −1 ) + θ ΘE ( at −13wt −1 )
=Φ γ 11 − θσ 2 + θ Θ ( Φ − Θ ) σ 2
و
γ 11 = E ( wt wt −11 ) = Φ γ 1 + Θθσ 2
;J.
; اB "
اH%.و
( Θ − Φ )2
γ 1 = −θσ 1 + 2
1 − Φ 2
Φ (Θ − Φ )
2
γ 11 = θσ Θ − Φ −
2
1 − Φ 2
(ت أنU إ3!2 4J26
~ اb'.و
γ 2 = γ 3 = ⋯ = γ 10 = 0
γ 13 = γ 11
γ k = Φ γ k −12 , k > 13
7CاA
اw. ا
;ا4
داL1
4J.
ت اB "
اV!L &و
1, k =0
θ
− , k =1
1+θ
2
0, k = 2,...,10
γ θ ( Θ − Φ )(1 − ΦΘ )
ρk = k = , k = 11
γ 0 1 + θ 2 1 + Θ2 − 2ΦΘ
( Θ − Φ )(1 − ΦΘ )
− , k = 12
1 + Θ2 − 2ΦΘ
ρ11 , k = 13
Φρ , k > 13
k −12
142
: 0 اذج اv ا9\ ا$دوال ا
ا
wt = (1 − ΘB s ) at SARIMA(0,d,0)(0,D,1)s ذج1!
-1
1, k =0
Θ
ρk = − , k=s
1+ Θ
2
0, otherwise
(1 − ΦB ) w
s
t = at SARIMA(0,d,0)(1,D,1)s ذج1!
-2
1, k =0
ks
ρ k = Φ , k = s, 2 s,...
0, otherwise
wt = (1 − θ B ) (1 − ΘB s ) at SARIMA(0,d,1)(0,D,1)s ذج1! -3
1, k =0
θ
− , k =1
1+θ
2
θΘ
, k = s −1
ρ k = (1 + θ )(1 + Θ )
2 2
Θ
− , k=s
1+ Θ
2
ρ s −1 , k = s +1
0, otherwise
(1 − ΦB ) w = (1 − ΘB ) a
s
t
s
t SARIMA(0,d,0)(1,D,1)s ذج1!
-4
1, k =0
( Θ − Φ )(1 − ΦΘ ) k s −1
ρk = − Φ , k = s, 2 s,...
1 + Θ − 2ΦΘ
2
0, otherwise
(1 − ΦB ) w = (1 − θ B ) a
s
t t SARIMA(0,d,1)(1,D,0)s ذج1!
-5
143
1, k =0
θ
− , k =1
1+θ
2
0, k = 2,..., s − 2
ρk = θ Φ
− , k = s −1
1+θ 2
Φ, k=s
ρ s −1 , k = s +1
Φ ρ k − s , k > s +1
wt = (1 − θ1B − θ 2 B 2 )(1 − ΘB12 ) at SARIMA(0,d,2)(0,D,1)s ذج1! -6
1, k =0
− θ1 (1 − θ 2 ) , k =1
1 + θ12 + θ 22
− θ2
, k =2
1 + θ12 + θ 22
θ 2Θ
, k = s−2
(1 + θ1 + θ 22 )(1 + Θ2 )
2
ρk =
θ1Θ (1 − θ 2 )
, k = s −1
(1 + θ 2 + θ 2 )(1 + Θ2 )
1 2
Θ
− , k=s
1+ Θ
2
ρ s −1 , k = s +1
ρ , k = s+2
s −2
0, otherwise
ء6 4!1!
ا4'&
!; ت ا7CاA
اw. ا
;ا4
&ت
ا1
"\ ا. ف
;"ض1
.O
3ة أ3N
: SARIMA(0,d,1)(1,D,0)12 ذج1!'
4
;
ل ا3ا
144
(1) H3
Φ = 0.6, θ = 0.5
A C F o f S A R I M A ( 0 ,d ,1 ) ( 1 ,D ,0 ) 1 2
C1 0 .5
0 .0
-0 .5
0 1 0 2 0 3 0 4 0 5 0
L a g
( 2) H3
Φ = 0.6, θ = −0.5
A C F o f S A R I M A (0 ,d ,1 ) ( 1 ,D ,0 ) 1 2
0 .7
0 .6
0 .5
0 .4
C1
0 .3
0 .2
0 .1
0 .0
0 1 0 2 0 3 0 4 0 5 0
L a g
(3) H3
Φ = −0.6, θ = 0.5
A C F o f S A R IM A (0 ,d ,1 )(1 ,D ,0 )1 2
0 .5
0 .0
C1
-0 .5
0 10 20 30 40 50
Lag
145
(4) H3
Φ = −0.6, θ = −0.5
A C F o f S A R IM A (0 ,d ,1 )(1 ,D ,0 )1 2
0 .5
C1
0 .0
-0 .5
0 1 0 20 3 0 4 0 5 0
L ag
دا ا ا \$ا9ا ا Vذج ا 0ا:Y O
& ا
" 4.1إ;Jق و bCأ !ط دا
4ا
;ا w.ا
Aا 7Cا
!'
78ذج ا
! 4!1ا
;`4b
و
O'3و H3K.م FNن أLاء ا
'!1ذج ا
! 4!1و zا
! 4!1وا
;A!'C 7ج ا
!; w1ا
!;%ك
&[C 76"Cات ا 4و&[Cات ' 4(Lا
;[bت ا
! 4!1وzا
! 4!1و 7Nا
'!ذج ا
;7
1%Cي إ %ار ذاFN 7Cن ا
;ا6.ت ا
Aا 4Cا
. cut off "6B 76"C 48
ا3ل ا
;
69 4ء 3Nة \".دوال ا
;ا w.ا
Aا 7Cا
\"(
78ا
'!ذج :
H3 -1دا
4ا
;ا w.ا
Aا 7Cا
1!'
78ذج wt = (1 − ΘB12 ) at
ا( Θ = 0.6
A C F o f S A R IM A (0 ,d ,0 )(0 ,D ,1 )1 2
0 .0
-0 .1
-0 .2
C1
-0 .3
-0 .4
-0 .5
-0 .6
0 1 0 2 0 3 0 4 0 5 0
L a g
Θ = −0.6 ب(
146
A C F o f S A R IM A (0 ,d ,0 )(0 ,D ,1 )1 2
0 .6
0 .5
0 .4
0 .3
C1
0 .2
0 .1
0 .0
-0 .1
-0 .2
0 1 0 2 0 3 0 4 0 5 0
L ag
(1 − ΦB ) w
12
t H3 -2دا
4ا
;ا w.ا
Aا 7Cا
1!'
78ذج = at
ا( Φ = 0.6
A C F o f S A R IM A (0 ,d ,1 )(0 ,D ,0 )1 2
0 .6
0 .5
0 .4
C1
0 .3
0 .2
0 .1
0 .0
0 1 0 2 0 3 0 4 0 5 0
L a g
ب( Φ = −0.6
A C F o f S A R IM A (0 ,d ,1 )(0 ,D ,0 )1 2
0 .0
-0 .1
-0 .2
C1
-0 .3
-0 .4
-0 .5
-0 .6
0 1 0 2 0 3 0 4 0 5 0
L a g
ﺃﻤﺜﻠﺔ :ﻝﻠﻤﺘﺴﻠﺴﻠﺔ ﺍﻝﺯﻤﻨﻴﺔ ﺍﻝﻤﻭﺴﻤﻴﺔ )ﻓﻲ ﺠﻤﻴﻊ ﺍﻷﻤﺜﻠﺔ ﺍﻝﺘﺎﻝﻴﺔ ﺇﻗﺭﺃ ﺴﻁﺭﺍ ﺒﺴﻁﺭ(
)z(t
56.3 55.7 55.8 56.3 57.2 59.1 71.5 72.2 72.7 61.5 57.4 56.9
55.3 54.9 54.9 54.9 54.6 57.7 68.2 70.6 71.0 60.0 56.0 54.4
53.3 52.8 53.0 53.4 54.3 58.2 67.4 71.0 69.8 59.4 55.6 54.6
147
53.4 53.0 53.0 53.2 54.2 58.0 67.5 70.1 68.2 56.6 54.9 54.0
52.9 52.6 52.8 53.0 53.6 56.1 66.1 69.8 69.3 61.2 57.5 54.9
53.4 52.7 53.0 52.9 55.4 58.7 67.9 70.0 68.7 59.3 56.4 54.5
52.8 52.8 53.2 55.3 55.8 58.2 65.3 67.9 68.3 61.7 56.4 53.9
52.6 52.1 52.4 51.6 52.7 57.3 65.1 71.5 69.9 61.9 57.3 55.1
53.6 53.4 53.5 53.3 53.9 52.7 61.0 69.9 70.4 59.4 56.3 54.3
53.5 53.0 53.2 52.5 53.4 56.5 65.3 70.7 66.9 58.2 55.3 53.4
52.1 51.5 51.5 52.4 53.3 55.5 64.2 69.6 69.3 58.5 55.3 53.6
52.3 51.5 51.7 51.5 52.2 57.1 63.6 68.8 68.9 60.1 55.6 53.9
53.3 53.1 53.5 53.5 53.9 57.1 64.7 69.4 70.3 62.6 57.9 55.8
54.8 54.2 54.6 54.3 54.8 58.1 68.1 73.3 75.5 66.4 60.5 57.7
55.8 54.7 55.0 55.6 56.4 60.6 70.8 76.4 74.8 62.2
1 ه4 ;!
اH3
7 0
z(t)
6 0
5 0
In d e x 5 0 1 0 0 1 5 0
7CاA
اw. ا
;ا4
دا
A u t o c o r r e la t io n F u n c t io n f o r z ( t )
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
2 12 22 32 42
Lag C o rr T LBQ Lag C o rr T LBQ Lag C o rr T LBQ Lag C o rr T LBQ
1 0 .7 7 1 0 .3 1 1 0 8 .1 0 13 0 .6 8 3 .3 4 6 9 4 .2 8 25 0 .6 1 2 .3 0 1 2 2 0 .0 7 37 0 .5 7 1 .8 2 1 7 0 9 .3 6
2 0 .3 3 2 .9 3 1 2 7 .3 4 14 0 .2 8 1 .3 0 7 0 9 .8 2 26 0 .2 5 0 .9 1 1 2 3 3 .1 8 38 0 .2 3 0 .7 3 1 7 2 1 .7 0
3 -0 .1 1 - 0 .9 9 1 2 9 .7 4 15 -0 .1 2 -0 .5 3 7 1 2 .4 3 27 -0 .1 1 -0 .4 2 1 2 3 5 .9 8 39 -0 .1 1 -0 .3 4 1 7 2 4 .4 3
4 -0 .3 8 - 3 .2 8 1 5 6 .8 3 16 -0 .3 6 -1 .6 5 7 3 8 .3 0 28 -0 .3 4 -1 .2 4 1 2 6 1 .2 9 40 -0 .3 3 -1 .0 2 1 7 4 9 .0 6
5 -0 .5 2 - 4 .2 2 2 0 7 .2 4 17 -0 .4 9 -2 .2 1 7 8 6 .2 8 29 -0 .4 7 -1 .6 7 1 3 0 8 .3 2 41 -0 .4 4 -1 .3 8 1 7 9 5 .0 0
6 -0 .5 6 - 4 .1 6 2 6 6 .5 5 18 -0 .5 3 -2 .3 3 8 4 3 .0 4 30 -0 .5 1 -1 .8 0 1 3 6 4 .7 2 42 -0 .4 8 -1 .4 8 1 8 4 9 .4 1
7 -0 .5 2 - 3 .5 0 3 1 6 .8 0 19 -0 .4 9 -2 .1 0 8 9 2 .3 6 31 -0 .4 8 -1 .6 5 1 4 1 4 .0 8 43 -0 .4 5 -1 .3 7 1 8 9 7 .0 5
8 -0 .3 8 - 2 .3 9 3 4 3 .7 1 20 -0 .3 7 -1 .5 4 9 2 0 .3 9 32 -0 .3 6 -1 .2 3 1 4 4 2 .7 1 44 -0 .3 4 -1 .0 4 1 9 2 5 .3 0
9 -0 .1 2 - 0 .7 1 3 4 6 .2 5 21 -0 .1 4 -0 .5 6 9 2 4 .2 4 33 -0 .1 4 -0 .4 9 1 4 4 7 .3 4
10 0 .2 9 1 .7 6 3 6 2 .0 8 22 0 .2 3 0 .9 5 9 3 5 .3 7 34 0 .1 9 0 .6 6 1 4 5 5 .7 9
11 0 .6 9 4 .1 5 4 5 3 .5 8 23 0 .6 1 2 .4 6 1 0 1 1 .3 1 35 0 .5 4 1 .8 2 1 5 2 0 .8 9
12 0 .8 8 4 .8 6 6 0 3 .8 7 24 0 .7 9 3 .1 0 1 1 4 0 .9 0 36 0 .7 2 2 .3 7 1 6 3 6 .4 9
78
ا7CاA
اw.وا
;ا
148
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r z ( t )
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
- 0 .2
- 0 .4
- 0 .6
- 0 .8
- 1 .0
2 12 22 32 42
Lag PAC T Lag P AC T Lag P AC T Lag P AC T
1 0 .7 7 1 0 .3 1 13 - 0 .4 2 - 5 .6 2 25 - 0 .1 3 - 1 .7 5 37 - 0 .0 6 - 0 .7 4
2 -0 .6 8 - 9 .0 1 14 0 .2 9 3 .8 9 26 - 0 .0 0 - 0 .0 5 38 - 0 .0 5 - 0 .6 2
3 -0 .0 6 - 0 .8 2 15 0 .0 1 0 .1 6 27 0 .0 4 0 .5 0 39 0 .0 1 0 .1 7
4 0 .0 1 0 .0 7 16 - 0 .0 7 - 0 .9 2 28 - 0 .1 0 - 1 .3 0 40 0 .0 0 0 .0 0
5 -0 .4 2 - 5 .5 5 17 0 .0 7 0 .9 5 29 - 0 .0 3 - 0 .3 4 41 - 0 .0 5 - 0 .6 3
6 -0 .1 8 - 2 .4 5 18 - 0 .0 1 - 0 .1 1 30 0 .0 1 0 .1 9 42 0 .0 5 0 .6 6
7 -0 .1 1 - 1 .4 8 19 - 0 .0 9 - 1 .1 5 31 0 .0 1 0 .0 7 43 - 0 .0 5 - 0 .6 4
8 -0 .2 2 - 2 .9 6 20 - 0 .0 5 - 0 .7 2 32 - 0 .0 4 - 0 .5 5 44 - 0 .0 3 - 0 .4 2
9 0 .1 8 2 .3 9 21 - 0 .0 8 - 1 .1 1 33 - 0 .0 4 - 0 .5 2
10 0 .5 7 7 .6 5 22 0 .1 0 1 .3 8 34 0 .0 1 0 .1 5
11 0 .2 6 3 .5 0 23 - 0 .0 3 - 0 .3 5 35 - 0 .0 2 - 0 .3 2
12 0 .1 6 2 .1 6 24 0 .0 3 0 .3 7 36 0 .0 7 0 .9 0
.4J.
ل ا3 ا7N 4%P وا4!1!
ا
!ط ا5
ﻤﺜﺎل ﺁﺨﺭ
z(t)
589 561 640 656 727 697 640 599 568 577 553 582
600 566 653 673 742 716 660 617 583 587 565 598
628 618 688 705 770 736 678 639 604 611 594 634
658 622 709 722 782 756 702 653 615 621 602 635
677 635 736 755 811 798 735 697 661 667 645 688
713 667 762 784 837 817 767 722 681 687 660 698
717 696 775 796 858 826 783 740 701 706 677 711
734 690 785 805 871 845 801 764 725 723 690 734
750 707 807 824 886 859 819 783 740 747 711 751
804 756 860 878 942 913 869 834 790 800 763 800
826 799 890 900 961 935 894 855 809 810 766 805
821 773 883 898 957 924 881 837 784 791 760 802
828 778 889 902 969 947 908 867 815 812 773 813
834 782 892 903 966 937 896 858 817 827 797 843
4 ;!
اH3
149
1000
900
z(t) 800
700
600
In d e x 50 100 150
7CاA
اw.ا
;ا
A u t o c o r r e la t io n F u n c t io n f o r z ( t )
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
2 12 22 32 42
L ag C o rr T LB Q Lag C o rr T LB Q Lag C o rr T LB Q Lag C o rr T LB Q
1 0 .8 9 1 1 .5 6 1 3 5 .9 4 1 3 0 .7 4 2 .9 6 9 5 4 .6 8 2 5 0 .5 8 1 .8 61 4 8 3 .1 1 3 7 0 .4 3 1 .2 71 7 8 4 .1 7
2 0 .7 8 6 .2 7 2 4 0 .1 3 1 4 0 .6 4 2 .4 11 0 3 0 .0 9 2 6 0 .4 9 1 .5 21 5 3 0 .9 9 3 8 0 .3 5 1 .0 01 8 1 0 .6 0
3 0 .6 2 4 .1 2 3 0 6 .7 2 1 5 0 .4 9 1 .7 91 0 7 4 .8 5 2 7 0 .3 5 1 .0 91 5 5 6 .4 5 3 9 0 .2 2 0 .6 41 8 2 1 .6 8
4 0 .4 9 2 .9 5 3 4 7 .9 7 1 6 0 .3 6 1 .3 11 0 9 9 .6 8 2 8 0 .2 4 0 .7 31 5 6 8 .2 0 4 0 0 .1 2 0 .3 31 8 2 4 .6 7
5 0 .4 3 2 .4 7 3 8 0 .0 9 1 7 0 .3 1 1 .0 91 1 1 7 .3 9 2 9 0 .1 9 0 .5 71 5 7 5 .4 4 4 1 0 .0 6 0 .1 81 8 2 5 .5 6
6 0 .3 8 2 .1 0 4 0 5 .0 2 1 8 0 .2 5 0 .9 01 1 2 9 .7 6 3 0 0 .1 4 0 .4 31 5 7 9 .5 5 4 2 0 .0 2 0 .0 51 8 2 5 .6 2
7 0 .4 1 2 .2 5 4 3 5 .5 4 1 9 0 .2 9 1 .0 11 1 4 5 .5 9 3 1 0 .1 7 0 .5 11 5 8 5 .5 2
8 0 .4 5 2 .4 0 4 7 2 .3 7 2 0 0 .3 2 1 .1 21 1 6 5 .4 2 3 2 0 .2 0 0 .6 01 5 9 3 .6 2
9 0 .5 6 2 .8 7 5 2 9 .0 7 2 1 0 .4 2 1 .4 41 1 9 9 .1 3 3 3 0 .2 8 0 .8 51 6 1 0 .3 5
10 0 .6 9 3 .3 4 6 1 4 .2 7 2 2 0 .5 3 1 .8 11 2 5 3 .8 1 3 4 0 .3 8 1 .1 51 6 4 1 .5 7
11 0 .7 7 3 .5 2 7 2 1 .7 2 2 3 0 .6 0 2 .0 31 3 2 5 .5 1 3 5 0 .4 5 1 .3 51 6 8 5 .2 7
12 0 .8 4 3 .6 1 8 5 2 .4 1 2 4 0 .6 7 2 .2 11 4 1 5 .2 9 3 6 0 .5 2 1 .5 31 7 4 2 .9 5
78
ا7CاA
اw.وا
;ا
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r z ( t )
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
2 12 22 32 42
Lag PAC T L ag PAC T L ag PAC T Lag PAC T
1 0 .8 9 1 1 .5 6 13 -0 .6 3 -8 .1 9 25 -0 .1 8 -2 .3 6 37 -0 .1 1 -1 .3 7
2 -0 .0 8 -1 .0 6 14 -0 .0 2 -0 .2 1 26 0 .0 8 1 .0 6 38 -0 .0 2 -0 .2 2
3 -0 .2 8 -3 .6 5 15 0 .0 7 0 .9 5 27 0 .0 6 0 .7 3 39 0 .0 4 0 .5 1
4 0 .0 3 0 .4 2 16 -0 .0 4 -0 .5 2 28 -0 .0 3 -0 .4 4 40 -0 .0 3 -0 .4 2
5 0 .3 5 4 .5 4 17 -0 .0 9 -1 .1 2 29 -0 .0 4 -0 .4 7 41 -0 .0 8 -1 .0 6
6 -0 .0 8 -1 .0 7 18 -0 .0 4 -0 .4 9 30 0 .0 0 0 .0 2 42 0 .0 1 0 .0 8
7 0 .2 8 3 .6 7 19 -0 .0 5 -0 .6 0 31 -0 .0 6 -0 .7 2
8 0 .0 9 1 .1 9 20 0 .0 3 0 .3 8 32 -0 .0 1 -0 .1 1
9 0 .4 0 5 .1 7 21 0 .0 4 0 .4 6 33 -0 .0 1 -0 .1 8
10 0 .3 0 3 .9 5 22 0 .0 5 0 .6 7 34 0 .0 3 0 .3 8
11 0 .0 6 0 .8 1 23 0 .0 5 0 .6 0 35 0 .0 0 0 .0 3
12 0 .2 2 2 .8 8 24 0 .0 5 0 .5 9 36 0 .0 1 0 .0 9
150
?z ل-
z(t)
302 262 218 175 100 077 043 047 049 069 152 205 246 294
242 181 107 056 049 047 047 071 151 244 280 230 185 148
098 061 046 045 055 048 115 185 276 220 181 151 083 055
049 042 046 074 103 200 237 247 215 182 080 046 065
040 044 063 085 185 247 231 167 117 079 045 040 038 041
069 152 232 282 255 161 107 053 040 039 034 035 056 097
210 260 257 210 125 080 042 035 031 032 050 092 189 256
250 198 136 073 039 032 030 031 045
H3K
اO
و
3 0 0
2 0 0
z(t)
1 0 0
0
In d e x 10 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 1 0 0
7CاA
اw. ا
;ا4
ودا
A u t o c o r r e la t io n F u n c t io n f o r z ( t )
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 1 5 2 5
L a g C o rr T L B Q L a g C o rr T L B Q L a g C o rr T L B Q L a g C o rr T L B Q
1 0 .8 1 8 .3 8 7 2 .2 2 8 -0 .4 3 -1 .7 8 3 1 0 .6 6 1 5 -0 .0 3 -0 .0 8 5 5 6 .1 8 2 2 0 .3 3 0 .9 2 7 5 6 .9 8
2 0 .4 3 2 .8 9 9 2 .3 2 9 -0 .0 4 -0 .1 8 3 1 0 .8 8 1 6 -0 .3 8 -1 .1 9 5 7 4 .0 7 2 3 0 .6 1 1 .7 1 8 0 8 .9 7
3 -0 .0 3 -0 .1 7 9 2 .4 0 1 0 0 .3 8 1 .5 4 3 2 8 .2 2 1 7 -0 .6 0 -1 .8 8 6 2 0 .2 2 2 4 0 .7 1 1 .9 3 8 7 9 .4 2
4 -0 .4 3 -2 .7 0 1 1 3 .1 3 1 1 0 .7 1 2 .8 2 3 8 9 .7 2 1 8 -0 .6 7 -2 .0 3 6 7 8 .3 6 2 5 0 .5 9 1 .5 5 9 2 8 .4 1
5 -0 .6 9 -4 .0 3 1 6 6 .3 7 1 2 0 .8 4 3 .1 0 4 7 6 .2 2 1 9 -0 .5 9 -1 .7 2 7 2 4 .1 7 2 6 0 .3 1 0 .8 0 9 4 2 .4 6
6 -0 .7 8 -3 .9 9 2 3 5 .1 7 1 3 0 .7 1 2 .4 2 5 3 8 .8 4 2 0 -0 .3 7 -1 .0 5 7 4 2 .2 1
7 -0 .6 8 -3 .0 9 2 8 9 .3 3 1 4 0 .3 7 1 .2 0 5 5 6 .1 0 2 1 -0 .0 4 -0 .1 0 7 4 2 .3 8
78
ا7CاA
اw. ا
;ا4
ودا
151
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r z ( t )
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 1 5 2 5
L a g P A C T L a g P A C T L a g P A C T L a g P A C T
1 0 .8 1 8 .3 8 8 -0 .0 3 -0 .3 1 1 5 0 .1 9 1 .9 3 2 2 0 .0 1 0 .1 5
2 -0 .7 0 -7 .1 7 9 0 .1 9 1 .9 1 1 6 0 .0 8 0 .8 4 2 3 0 .0 7 0 .7 0
3 -0 .2 9 -2 .9 8 1 0 0 .2 7 2 .7 6 1 7 0 .1 1 1 .1 0 2 4 -0 .0 7 -0 .7 2
4 -0 .2 2 -2 .2 8 1 1 0 .1 8 1 .8 2 1 8 0 .0 4 0 .3 9 2 5 -0 .0 2 -0 .2 3
5 -0 .1 6 -1 .6 1 1 2 0 .0 7 0 .7 4 1 9 -0 .0 2 -0 .2 2 2 6 0 .1 0 1 .0 1
6 -0 .3 2 -3 .2 9 1 3 -0 .1 0 -0 .9 9 2 0 -0 .0 3 -0 .2 9
7 -0 .1 6 -1 .6 0 1 4 -0 .1 5 -1 .5 2 2 1 0 .0 4 0 .4 6
.4%P وا4!1& (ي ا !طC @ ا
!; تA هHوآ
: Y O ا0 ت ا
ا,ذج ا7 v 4 إ{ ق دوال
ن =قFN ARIMA & !ذج4+ 4
5 7 ه4!1!
ا4'&
! ان !ذج ا
!; ت ا.
ذج1!'
&"
) ا2JCذج و1!'
اH3 7 ا
;"فQ5 & 4J.
ق ا6
~ اb 7 هO"& H&";
ا
z
'!ذجJ. در'ه7;
ت اXق وا
!"د6
اV!L .:(';
) اU & و4 %b;
;(رات ا9وا
.wJN ^P1;
("\ ا
'!ذج:(';
; دوال اK ف1 .'( ه6'C 4!1!
ا
: SARIMA(0,0,0)(0,1,1)12 ذج1!'
:(';
ا4
دا-1
H3K
ذج ا1!'
اW;32و
(1 − B ) z = (1 − ΘB ) a
12
t
12
t
4Bوb
ا4
& ا
!"د
zn +ℓ = zn +ℓ−12 + an +ℓ − Θan + ℓ−12
7
;
ات آ:(';
ل ا1 %
ا3!2
zn (1) = zn −11 − Θan −11
zn ( 2 ) = zn −10 − Θan −10
⋮
zn (12 ) = zn − Θan
zn ( ℓ ) = zn ( ℓ − 12 ) , ℓ ≥ 12
أو
zn +ℓ −12 − Θan + ℓ−12 , ℓ = 1, 2,...,12
zn ( ℓ ) =
zn ( ℓ − 12 ) , ℓ > 12
^ أنPوا
152
zn (1) = zn (13) = zn ( 25 ) = ⋯
zn ( 2 ) = zn (14 ) = zn ( 26 ) = ⋯
⋮
zn (12 ) = zn ( 24 ) = zn ( 36 ) = ⋯
:(';
ء ا6 أ2(C
V en ( ℓ ) = σ 2 (1 + ψ 12 + ⋯ + ψ ℓ2−1 )
(f
ه ذ.) 4B "
. 6"C اوزان4
ودا
1 − Θ, j = 12, 24,...
ψj =
0, otherwise
:(';
ء ا6 أ2(C 4Y+ 7N \ اوزان21";.و
ℓ − 1 2
V en ( ℓ ) = σ 2 1 + (1 − Θ )
12
. x & ^%
ا
ء ا7'"C x Q5
: SARIMA(0,1,1)(0,1,1)12 ذج1!'
:(';
ا4
دا-2
H3K
ذج ا1!'
اW;32و
(1 − B ) (1 − B12 ) zt = (1 − θ B ) (1 − ΘB12 ) at
4Bوb
ا4
& ا
!"د
zn +ℓ = zn +ℓ −1 + zn +ℓ −12 − zn +ℓ −13 + an + ℓ − θ an +ℓ −1 − Θan + ℓ−12 + θ Θan + ℓ−13
7
;
ات آ:(';
ل ا1 %
ا3!2
zn (1) = zn + zn −11 − zn −13 − θ an − Θan −11 + θ Θan −12
zn ( 2 ) = zn (1) + zn −10 − zn −11 − Θan −10 + θ Θan −11
⋮
zn (12 ) = zn (11) + zn − zn −1 − Θan + θ Θan −1
zn (13) = zn (12 ) + zn (1) − zn + θ Θan
zn ( ℓ ) = zn ( ℓ − 1) + zn ( ℓ − 12 ) − zn ( ℓ − 13)
4
) أوJ. اA3وه
153
zn (1) = zn + zn −11 − zn −13 − θ an − Θan −11 + θ Θan −12
zn ( 2 ) = zn (1) + zn −10 − zn −11 − Θan −10 + θ Θan −11
⋮
zn (12 ) = zn (11) + zn − zn −1 − Θan + θ Θan −1
zn (13) = zn (12 ) + zn (1) − zn + θ Θan
42ار3C 4B و
zn ( ℓ ) = zn ( ℓ − 1) + zn ( ℓ − 12 ) − zn ( ℓ − 13) , ℓ > 13
.ات:(';
ب & ا16!
ا
"د ا1C 3!2
: 0 ت ا
ا, اv 0ت دراrK و-
أ
:4
;
هات اK!
( ا6'2 SARIMA 48 & ذج1!
د2ول إ%
ف1 : (1) ل-
z(t)
589 561 640 656 727 697 640 599 568 577 553 582
600 566 653 673 742 716 660 617 583 587 565 598
628 618 688 705 770 736 678 639 604 611 594 634
658 622 709 722 782 756 702 653 615 621 602 635
677 635 736 755 811 798 735 697 661 667 645
688 713 667 762 784 837 817 767 722 681 687 660
698 717 696 775 796 858 826 783 740 701 706 677
711 734 690 785 805 871 845 801 764 725 723 690
734 750 707 807 824 886 859 819 783 740 747 711
751 804 756 860 878 942 913 869 834 790 800 763
800 826 799 890 900 961 935 894 855 809 810 766
805 821 773 883 898 957 924 881 837 784 791 760
802 828 778 889 902 969 947 908 867 815 812 773
813 834 782 892 903 966 937 896 858 817 827 797
843
هاتK!
7'&
اw6[!
وا
154
1 00 0
90 0
80 0
z(t)
70 0
60 0
In d e x 5 0 1 00 15 0
7!Uرz1
H21%;. X أو2(;
اS(]
f
A
w1;!
وا2(;
ا7N ةJ; & z 4 ;!
ان ا5 2
O
7'&
اw6[!
و
) اyt = ln ( zt ) أي
6 .9
6 .8
6 .7
y(t)
6 .6
6 .5
6 .4
6 .3
In d e x 5 0 1 0 0 1 5 0
قb
اA{
f
A
w1;!
ا7N ةJ; & z الCX 3
و2(;
ا7N تJ; ا4 ;!
ان ا5
7
;
اH3K
اO
وxt = (1 − B ) yt = (1 − B ) ln ( zt ) اول
0 .1 5
0 .1 0
0 .0 5
y(t)-y(t-1)
0 .0 0
-0 .0 5
-0 .1 0
In d e x 5 0 1 0 0 1 5 0
w. وا
;ا7CاA
اw. إ
دوال ا
;اI''
.w1;!
وا2(;
& اH آ7N ةJ; & نs ا4 ;!
ا
O
78
ا7CاA
ا
155
A u t o c o r r e la t io n F u n c t io n f o r y ( t ) - y ( t
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
10 20 30 40
L ag C o rr T LB Q Lag C o rr T LBQ L ag C o rr T LBQ L ag C o rr T LBQ
1 0 .0 1 0 .1 2 0 .0 2 13 0 .0 2 0 .1 2 2 6 1 .4 7 25 0 .0 2 0 .0 8 4 9 7 .3 5 3 7 0 .0 1 0 .0 5 7 1 3 .1 6
2 0 .2 5 3 .2 3 1 0 .7 0 14 0 .2 3 1 .5 1 2 7 1 .2 9 26 0 .2 1 1 .1 0 5 0 6 .5 0 3 8 0 .2 0 0 .9 0 7 2 2 .0 6
3 - 0 .0 8 -1 .0 1 1 1 .8 8 15 - 0 .0 7 - 0 .4 7 2 7 2 .3 0 27 -0 .0 6 - 0 .2 9 5 0 7 .1 6 3 9 - 0 .0 4 -0 .1 9 7 2 2 .4 5
4 - 0 .3 7 -4 .4 8 3 5 .6 0 16 - 0 .3 4 - 2 .1 8 2 9 3 .7 3 28 -0 .3 1 - 1 .5 7 5 2 6 .4 5 4 0 - 0 .2 7 -1 .2 1 7 3 8 .8 6
5 - 0 .0 6 -0 .6 8 3 6 .2 8 17 - 0 .0 6 - 0 .3 9 2 9 4 .4 5 29 -0 .0 6 - 0 .2 9 5 2 7 .1 4 4 1 - 0 .0 5 -0 .2 4 7 3 9 .5 1
6 - 0 .5 0 -5 .4 1 7 9 .9 7 18 - 0 .4 6 - 2 .8 8 3 3 4 .6 7 30 -0 .4 2 - 2 .0 9 5 6 3 .0 5
7 - 0 .0 4 -0 .3 3 8 0 .1 9 19 - 0 .0 3 - 0 .2 0 3 3 4 .8 8 31 -0 .0 2 - 0 .1 2 5 6 3 .1 6
8 - 0 .3 5 -3 .3 0 1 0 2 .3 7 20 - 0 .3 2 - 1 .9 3 3 5 4 .9 6 32 -0 .3 0 - 1 .4 7 5 8 2 .0 5
9 - 0 .0 5 -0 .4 3 1 0 2 .8 0 21 - 0 .0 5 - 0 .2 6 3 5 5 .3 6 33 -0 .0 5 - 0 .2 4 5 8 2 .5 8
10 0 .2 3 2 .0 1 1 1 2 .2 9 22 0 .2 1 1 .2 0 3 6 3 .5 5 34 0 .1 8 0 .8 9 5 8 9 .8 4
11 0 .0 1 0 .1 1 1 1 2 .3 2 23 0 .0 1 0 .0 8 3 6 3 .5 8 35 0 .0 0 0 .0 2 5 8 9 .8 5
12 0 .9 0 7 .7 4 2 6 1 .4 1 24 0 .8 2 4 .7 6 4 9 7 .3 0 36 0 .7 6 3 .6 3 7 1 3 .1 4
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r y ( t ) - y ( t
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
- 0 .2
- 0 .4
- 0 .6
- 0 .8
- 1 .0
10 20 30 40
L ag PAC T L ag PAC T L ag PAC T L ag PAC T
1 0 .0 1 0 .1 2 13 - 0 .0 4 - 0 .4 8 25 -0 .1 3 - 1 .6 7 37 -0 .0 4 -0 .5 1
2 0 .2 5 3 .2 3 14 - 0 .3 3 - 4 .2 6 26 -0 .0 5 - 0 .6 1 38 0 .0 2 0 .3 0
3 -0 .0 9 - 1 .2 0 15 0 .0 0 0 .0 6 27 0 .0 6 0 .8 4 39 0 .0 4 0 .4 8
4 -0 .4 6 - 5 .9 7 16 0 .1 8 2 .2 9 28 0 .0 0 0 .0 5 40 0 .0 6 0 .8 1
5 -0 .0 2 - 0 .2 2 17 0 .0 1 0 .0 7 29 -0 .0 7 - 0 .9 4 41 -0 .0 2 -0 .2 0
6 -0 .3 5 - 4 .5 2 18 0 .0 8 1 .0 6 30 0 .0 3 0 .3 4
7 -0 .1 4 - 1 .8 2 19 - 0 .0 8 - 1 .0 4 31 0 .0 7 0 .9 3
8 -0 .4 8 - 6 .2 6 20 0 .0 8 0 .9 8 32 -0 .0 1 - 0 .0 9
9 -0 .4 1 - 5 .3 6 21 0 .0 2 0 .2 9 33 -0 .1 0 - 1 .3 1
10 -0 .2 3 - 2 .9 9 22 - 0 .0 4 - 0 .5 3 34 0 .0 6 0 .8 0
11 -0 .5 8 - 7 .5 5 23 0 .0 2 0 .2 9 35 0 .0 3 0 .3 9
12 0 .6 3 8 .0 8 24 0 .0 8 1 .0 4 36 0 .0 1 0 .1 3
و12 تb[;
' اO!B ! ن1& ةJ; & z 4 ;!
ان ا7CاA
اw. ا
;ا4
& دا5 2
wt = (1 − B12 ) (1 − B ) ln ( zt ) اول7!1!
ق اb
اA{
f
A
ءw(. &[;C 36 و24
2b;
ا اA" ه. O!
و
0 .0 5
y(t)-y(t-1)12
0 .0 0
-0 .0 5
In d e x 50 100 150
O
78
ا7CاA
اw. وا
;ا7CاA
اw. دوال ا
;اL1
156
A u t o c o r r e la t io n F u n c t io n f o r y ( t ) - y ( t
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 15 25 35
Lag C o rr T LBQ Lag C o rr T LBQ Lag C o rr T LBQ L ag C o rr T LBQ
1 - 0 .2 1 -2 .6 5 7 .1 5 13 0 .1 8 1 .8 0 5 6 .6 4 25 0 .0 7 0 .6 2 6 7 .6 7 3 7 - 0 .0 6 - 0 .5 4 7 3 .7 1
2 - 0 .0 1 -0 .1 3 7 .1 7 14 -0 .0 7 -0 .7 0 5 7 .5 4 26 - 0 .0 0 - 0 .0 2 6 7 .6 7 3 8 0 .0 3 0 .2 6 7 3 .8 8
3 0 .1 0 1 .1 4 8 .6 3 15 -0 .0 5 -0 .4 8 5 7 .9 7 27 - 0 .0 6 - 0 .5 7 6 8 .3 8
4 - 0 .1 3 -1 .4 9 1 1 .2 0 16 0 .0 3 0 .2 6 5 8 .1 0 28 0 .0 3 0 .3 0 6 8 .5 8
5 - 0 .1 0 -1 .1 1 1 2 .6 7 17 0 .1 2 1 .1 3 6 0 .5 8 29 - 0 .1 0 - 0 .9 0 7 0 .4 0
6 - 0 .0 2 -0 .2 8 1 2 .7 7 18 -0 .0 0 -0 .0 2 6 0 .5 8 30 0 .0 1 0 .0 8 7 0 .4 1
7 0 .1 2 1 .3 3 1 4 .9 5 19 -0 .1 1 -1 .0 8 6 2 .9 1 31 0 .0 4 0 .3 8 7 0 .7 4
8 0 .0 5 0 .5 4 1 5 .3 1 20 0 .0 3 0 .2 5 6 3 .0 4 32 0 .0 0 0 .0 4 7 0 .7 4
9 - 0 .0 5 -0 .5 8 1 5 .7 5 21 -0 .0 2 -0 .2 0 6 3 .1 2 33 0 .0 2 0 .1 9 7 0 .8 3
10 0 .1 3 1 .4 5 1 8 .4 9 22 -0 .0 9 -0 .8 0 6 4 .4 5 34 0 .0 0 0 .0 3 7 0 .8 3
11 - 0 .0 1 -0 .0 9 1 8 .5 0 23 0 .1 1 0 .9 9 6 6 .5 2 35 0 .0 9 0 .8 0 7 2 .3 6
12 - 0 .4 4 -4 .9 0 5 0 .8 8 24 -0 .0 4 -0 .3 8 6 6 .8 2 36 - 0 .0 6 - 0 .5 1 7 2 .9 9
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r y ( t ) - y ( t
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
- 0 .2
- 0 .4
- 0 .6
- 0 .8
- 1 .0
5 15 25 35
L ag PAC T L ag PAC T L ag PAC T L ag PAC T
1 -0 .2 1 - 2 .6 5 13 - 0 .0 0 - 0 .0 3 25 0 .1 2 1 .5 1 37 0 .1 0 1 .2 3
2 -0 .0 6 - 0 .7 3 14 - 0 .0 2 - 0 .2 5 26 -0 .0 1 - 0 .1 5 38 -0 .0 6 -0 .7 9
3 0 .0 8 1 .0 6 15 0 .0 0 0 .0 0 27 -0 .1 2 - 1 .4 4
4 -0 .0 9 - 1 .1 6 16 - 0 .1 1 - 1 .4 0 28 -0 .0 7 - 0 .9 1
5 -0 .1 5 - 1 .8 2 17 0 .0 4 0 .4 7 29 0 .0 4 0 .4 4
6 -0 .1 0 - 1 .1 9 18 0 .0 1 0 .1 2 30 -0 .0 4 - 0 .5 3
7 0 .1 1 1 .4 1 19 - 0 .0 4 - 0 .5 2 31 -0 .0 5 - 0 .6 0
8 0 .1 1 1 .4 2 20 - 0 .0 3 - 0 .3 5 32 0 .0 2 0 .1 9
9 -0 .0 4 - 0 .5 1 21 - 0 .0 3 - 0 .3 1 33 -0 .0 1 - 0 .1 2
10 0 .0 6 0 .8 0 22 0 .0 3 0 .3 2 34 -0 .0 2 - 0 .1 9
11 0 .0 5 0 .5 9 23 0 .0 7 0 .9 1 35 0 .1 8 2 .2 1
12 -0 .4 2 - 5 .2 3 24 - 0 .2 7 - 3 .4 1 36 -0 .2 1 - 2 .5 7
wt = (1 − B12 ) (1 − B ) ln ( zt ) 4 ;!
78
ا7CاA
اw. وا
;ا7CاA
اw.& ا
!ط دوال ا
;ا
ذج1!'
( ا6
اي0 و0 7 ه4'3!!
اq وp )B
ان
(1 − B ) (1 − B ) ln ( z ) = (1 − ΘB ) a
12
t
12
t
ذج1!'
ا اA( ه62 MINITAB 7N 7
;
ا& اSARIMA(0,1,0)(0,1,1)12 1ه
ARIMA 0 1 0 0 1 1 12 'y(t)' ;
NoConstant.
zt = e yt H21%;
ي ا48O'
_ ا8;'
ل ا1 %
وyt = ln ( zt ) '&[; ا
' ا5X
:f ا
MTB > Name c14 = 'RESI3' c15 = 'FITS3'
MTB > ARIMA 0 1 0 0 1 1 12 'y(t)' 'RESI3' 'FITS3';
SUBC> NoConstant;
SUBC> Forecast 24 c7 c8 c9;
157
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot.
ARIMA Model
ARIMA model for y(t)
Estimates at each iteration
Iteration SSE Parameters
0 0.0228597 0.100
1 0.0204943 0.250
2 0.0187066 0.400
3 0.0174234 0.550
4 0.0169841 0.684
5 0.0169841 0.683
6 0.0169841 0.683
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
SMA 12 0.6831 0.0610 11.20
Differencing: 1 regular, 1 seasonal of order 12
Number of observations: Original series 168, after
differencing 155
Residuals: SS = 0.0165799 (backforecasts excluded)
MS = 0.0001077 DF = 154
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 9.0(DF=11) 29.9(DF=23) 44.5(DF=35)
59.4(DF=47)
Forecasts from period 168
95 Percent Limits
158
Period Forecast Lower Upper
Actual
169 6.76750 6.74716 6.78784
170 6.70901 6.68024 6.73778
171 6.83815 6.80292 6.87338
172 6.85381 6.81313 6.89450
173 6.92288 6.87739 6.96836
174 6.89349 6.84366 6.94331
175 6.84654 6.79272 6.90035
176 6.80008 6.74255 6.85761
177 6.74395 6.68293 6.80497
178 6.75028 6.68596 6.81461
179 6.70664 6.63918 6.77410
180 6.75999 6.68952 6.83045
181 6.79052 6.71514 6.86590
182 6.73203 6.65203 6.81203
183 6.86117 6.77680 6.94554
184 6.87684 6.78832 6.96535
185 6.94590 6.85342 7.03838
186 6.91651 6.82023 7.01279
187 6.86956 6.76962 6.96950
188 6.82310 6.71963 6.92657
189 6.76697 6.66009 6.87385
190 6.77330 6.66312 6.88349
191 6.72966 6.61627 6.84305
192 6.78301 6.66649 6.89952
1 ه4 ;!
@ اAO
;حJ!
ذج ا1!'
أي أن ا
(1 − B ) (1 − B ) ln ( z ) = (1 − 0.683B ) a ,
12
t
12
t at ∼ N ( 0, 0.0001077 )
ان5X
Θ = 0.683, s.e. ( Θ ) = 0.061, t = 11.2
.421'"!
ا7
)"!
أي ان ا
:I> ا ا/a
\0 إ?ر ا
MTB > ZTest 0.0 0.0103778 'RESI3';
159
SUBC> Alternative 0.
Z-Test
Test of mu = 0.000000 vs mu not = 0.000000
The assumed sigma = 0.0104
Variable N Mean StDev SE Mean Z P
RESI3 155 -0.000111 0.010375 0.000834 -0.13 0.89
اb+ 7Bا1(
اw1;& \ أنN X اي0.05 & ( اآ7 وهP-value=0.89 ان ا
ـ5X
Iإ?ر ا ا ا
MTB > Runs 0 'RESI3'.
Runs Test
RESI3
K = 0.0000
The observed number of runs = 70
The expected number of runs = 78.1097
72 Observations above K 83 below
The test is significant at 0.1893
Cannot reject at alpha = 0.05
7Bا1(
ا48ا1K 4PN \N X ' اي ا0.1893 ' ي1'"& ;(ر9ا
:Iل ا ا, 0إ?ر إ
78
ا7CاA
اw. وا
;ا7CاA
اw.دوال ا
;ا
160
A C F o f R e s id u a ls f o r y ( t )
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e a u t o c o r r e l a t io n s )
1 .0
0 .8
0 .6
Autocorrelation 0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
3 6 9 12 15 18 21 24 27 30 33 36 39
Lag
P A C F o f R e s id u a ls f o r y ( t )
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e p a r t ia l a u t o c o r r e l a t io n s )
1 .0
0 .8
Partial Autocorrelation
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
3 6 9 12 15 18 21 24 27 30 33 36 39
Lag
.4J; & 7ON 4"(= S وإذا آ46. &;اz O ا
(`ء أي ا4 `
ا !ط ا76"C O ا5
:I ا اi إ?ر
161
Histogram of the Residuals
(response is y(t))
30
Frequency
20
10
-0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04
Residual
Normal Probability Plot of the Residuals
(response is y(t))
0.04
0.03
0.02
Residual
0.01
0.00
-0.01
-0.02
-0.03
-3 -2 -1 0 1 2 3
Normal Score
K-S test for Residuals
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-0.03 -0.02 -0.01 0.00 0.01 0.02 0.03
RESI3
Average: -0.0001115 Kolmogorov-Smirnov Normality Test
StDev: 0.0103754 D+: 0.074 D-: 0.045 D : 0.074
N: 155 Approximate P-Value: 0.041
' ي1'"& ;(ر9 اذا ا0.05 & HB ا7 وه0.041 76"2 K-S ;(ر9 P-value ان ا
ـ5X
.7Bا1(
ا4"(= 4PN \N
X ايα = 0.05
162
: ام ا ذج:0c$ 4ا
)
. O!
و:('C ;اتN 95% V& 4(J; & 4!B 24 :(';
. '!B 4J.
ت اL[!
ا7N
:7
;
ا
1150
1050
Forecast
950
850
750
0 5 10 15 20 25
T im e
:(';
ود ا5ات و:(';
اV& 4 ;!
7
;
وا
) ا
1150
1050
950
Forecast
850
750
650
550
0 100 200
Time
: 0 دراK
:4
;
هات اK!
( ا6'2 SARIMA 48 & ذج1!
د2ول إ%
ف1
z(t)
56.3 55.7 55.8 56.3 57.2 59.1 71.5 72.2
72.7 61.5 57.4 56.9 55.3 54.9 54.9 54.9
54.6 57.7 68.2 70.6 71.0 60.0 56.0 54.4
163
53.3 52.8 53.0 53.4 54.3 58.2 67.4 71.0
69.8 59.4 55.6 54.6 53.4 53.0 53.0 53.2
54.2 58.0 67.5 70.1 68.2 56.6 54.9 54.0
52.9 52.6 52.8 53.0 53.6 56.1 66.1 69.8
69.3 61.2 57.5 54.9 53.4 52.7 53.0 52.9
55.4 58.7 67.9 70.0 68.7 59.3 56.4 54.5
52.8 52.8 53.2 55.3 55.8 58.2 65.3 67.9
68.3 61.7 56.4 53.9 52.6 52.1 52.4 51.6
52.7 57.3 65.1 71.5 69.9 61.9 57.3 55.1
53.6 53.4 53.5 53.3 53.9 52.7 61.0 69.9
70.4 59.4 56.3 54.3 53.5 53.0 53.2 52.5
53.4 56.5 65.3 70.7 66.9 58.2 55.3 53.4
52.1 51.5 51.5 52.4 53.3 55.5 64.2 69.6
69.3 58.5 55.3 53.6 52.3 51.5 51.7 51.5
52.2 57.1 63.6 68.8 68.9 60.1 55.6 53.9
53.3 53.1 53.5 53.5 53.9 57.1 64.7 69.4
70.3 62.6 57.9 55.8 54.8 54.2 54.6 54.3
54.8 58.1 68.1 73.3 75.5 66.4 60.5 57.7
55.8 54.7 55.0 55.6 56.4 60.6 70.8 76.4
74.8 62.2
7'&
اw6[!
ا
7 0
z(t)
6 0
5 0
In d e x 5 0 1 0 0 1 5 0
wt = (1 − B ) zt w1;!
ار اJ;9 ق اولb
اA{.
164
1 0
y(t)
0
-1 0
In d e x 5 0 1 0 0 1 5 0
78L 7C ذاw.اC و7C ذاw.اC دوالO
و
Autocorrelation Function for y(t)
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 12 22 32 42
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 0.47 6.32 40.60 13 0.43 2.76 354.69 25 0.41 2.06 638.60 37 0.40 1.72 896.48
2 -0.02 -0.22 40.67 14 -0.01 -0.03 354.69 26 -0.00 -0.01 638.60 38 0.02 0.07 896.54
3 -0.37 -4.10 65.83 15 -0.33 -2.00 375.77 27 -0.29 -1.43 656.89 39 -0.27 -1.14 913.29
4 -0.28 -2.82 80.10 16 -0.25 -1.49 388.04 28 -0.23 -1.08 667.77 40 -0.21 -0.89 923.75
5 -0.21 -2.00 87.92 17 -0.19 -1.12 395.15 29 -0.17 -0.82 674.14 41 -0.17 -0.71 930.47
6 -0.19 -1.77 94.41 18 -0.17 -0.97 400.68 30 -0.16 -0.78 679.93 42 -0.15 -0.61 935.55
7 -0.21 -1.92 102.29 19 -0.19 -1.09 407.71 31 -0.17 -0.83 686.56 43 -0.16 -0.65 941.37
8 -0.27 -2.46 115.89 20 -0.25 -1.42 419.89 32 -0.23 -1.06 697.63 44 -0.20 -0.83 950.82
9 -0.32 -2.85 135.47 21 -0.31 -1.76 439.21 33 -0.28 -1.31 714.89
10 -0.00 -0.02 135.48 22 -0.01 -0.06 439.24 34 -0.01 -0.06 714.93
11 0.46 3.91 176.37 23 0.42 2.35 475.10 35 0.37 1.70 745.11
12 0.86 6.71 318.37 24 0.79 4.28 602.75 36 0.72 3.27 860.42
Partial Autocorrelation Function for y(t)
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 12 22 32 42
Lag PAC T Lag PAC T Lag PAC T Lag PAC T
1 0.47 6.32 13 -0.27 -3.62 25 -0.02 -0.33 37 0.03 0.38
2 -0.32 -4.21 14 -0.01 -0.17 26 -0.06 -0.74 38 -0.03 -0.35
3 -0.30 -4.00 15 0.01 0.13 27 0.07 0.99 39 -0.01 -0.17
4 0.09 1.22 16 -0.07 -0.95 28 -0.00 -0.05 40 0.02 0.23
5 -0.23 -3.05 17 -0.01 -0.16 29 -0.02 -0.26 41 -0.04 -0.56
6 -0.25 -3.33 18 0.02 0.24 30 -0.04 -0.60 42 0.03 0.35
7 -0.16 -2.09 19 0.03 0.44 31 0.01 0.20 43 -0.00 -0.00
8 -0.43 -5.67 20 0.07 0.89 32 0.03 0.37 44 0.03 0.43
9 -0.64 -8.56 21 -0.12 -1.62 33 -0.02 -0.25
10 -0.31 -4.11 22 -0.02 -0.26 34 0.00 0.06
11 -0.16 -2.18 23 -0.06 -0.86 35 -0.09 -1.17
12 0.41 5.42 24 0.13 1.74 36 0.05 0.61
4 ;!
';_ ا2 وwt = (1 − B12 ) (1 − B ) zt أي12 4(C
& ا7!1& 2bC
;ج إ%C O
ى ا
4
;
ا
165
5
w(t) 0
-5
In d e x 50 100 150
78L 7C ذاw.اC و7C ذاw.اC دوالO
و
A u t o c o r r e la t io n F u n c t io n f o r w ( t )
1 .0
Autocorrelation
0 .8
0 .6
0 .4
0 .2
0 .0
- 0 .2
- 0 .4
- 0 .6
- 0 .8
- 1 .0
10 20 30 40
L ag C o rr T LBQ L ag C o rr T LBQ L ag C o rr T LBQ Lag C o rr T LBQ
1 -0 .0 1 -0 .1 9 0 .0 4 13 -0 . 0 8 -0 .8 4 6 0 .6 9 25 -0 .0 8 -0 .7 6 8 3 .0 9 3 7 0 .0 5 0 .4 6 1 2 2 .4 0
2 -0 .1 9 -2 .4 8 6 .3 1 14 0 .1 0 0 .9 8 6 2 .5 0 26 -0 .1 3 -1 .1 8 8 6 .2 8 3 8 0 .1 0 0 .8 4 1 2 4 .5 2
3 -0 .2 4 -2 .9 7 1 6 .1 0 15 0 .1 2 1 .1 5 6 5 .0 1 27 -0 .0 1 -0 .1 3 8 6 .3 3 3 9 0 .0 2 0 .1 5 1 2 4 .5 9
4 -0 .0 0 -0 .0 3 1 6 .1 0 16 -0 . 0 2 -0 .1 7 6 5 .0 7 28 0 .0 6 0 .5 6 8 7 .0 7 4 0 -0 .0 2 -0 .1 6 1 2 4 .6 7
5 0 .0 0 0 .0 3 1 6 .1 0 17 -0 . 0 3 -0 .2 8 6 5 .2 2 29 0 .1 2 1 .0 9 8 9 .9 4 4 1 -0 .1 2 -1 .0 1 1 2 7 .8 4
6 -0 .0 4 -0 .5 0 1 6 .4 1 18 0 .1 0 1 .0 1 6 7 .2 4 30 -0 .1 5 -1 .3 6 9 4 .4 7
7 -0 .0 4 -0 .4 9 1 6 .7 1 19 0 .0 7 0 .7 0 6 8 .2 3 31 -0 .0 8 -0 .7 2 9 5 .8 0
8 -0 .0 3 -0 .3 1 1 6 .8 3 20 -0 . 0 1 -0 .0 5 6 8 .2 4 32 -0 .0 3 -0 .2 8 9 6 .0 1
9 0 .2 2 2 .5 6 2 5 .2 6 21 -0 . 2 0 -1 .9 4 7 5 .9 4 33 0 .1 9 1 .6 6 1 0 3 .2 0
10 0 .1 0 1 .1 6 2 7 .1 3 22 -0 . 0 2 -0 .2 3 7 6 .0 5 34 0 .1 7 1 .4 9 1 0 9 .1 9
11 0 .0 8 0 .9 3 2 8 .3 7 23 0 .1 0 0 .9 5 7 8 .0 2 35 -0 .1 0 -0 .8 6 1 1 1 .2 4
12 -0 .4 2 -4 .6 3 5 9 .4 0 24 0 .1 4 1 .3 0 8 1 .7 8 36 -0 .2 2 -1 .9 2 1 2 1 .7 7
P a r t ia l A u t o c o r r e la t io n F u n c t io n f o r w ( t )
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
10 20 30 40
L ag PAC T L ag PAC T L ag PAC T Lag PAC T
1 -0 .0 1 -0 .1 9 13 -0 .0 5 -0 .6 9 25 -0 .1 2 -1 .4 9 3 7 0 .0 2 0 .2 4
2 -0 .1 9 -2 .4 8 14 0 .0 1 0 .1 5 26 -0 .0 8 -0 .9 7 3 8 -0 .0 8 - 1 .0 6
3 -0 .2 6 -3 .2 8 15 -0 .0 4 -0 .5 0 27 -0 .0 3 -0 .3 2 3 9 0 .0 3 0 .4 4
4 -0 .0 7 -0 .8 5 16 -0 .0 3 -0 .3 3 28 -0 .0 3 -0 .3 6 4 0 0 .0 5 0 .6 3
5 -0 .1 1 -1 .4 3 17 -0 .0 2 -0 .2 3 29 0 .0 7 0 .8 7 4 1 -0 .0 3 - 0 .3 6
6 -0 .1 4 -1 .8 4 18 0 .0 6 0 .7 8 30 -0 .1 1 -1 .3 9
7 -0 .1 1 -1 .4 6 19 0 .0 5 0 .6 2 31 -0 .0 2 -0 .2 9
8 -0 .1 3 -1 .6 6 20 -0 .0 3 -0 .4 2 32 -0 .1 0 -1 .2 9
9 0 .1 4 1 .7 6 21 -0 .0 5 -0 .6 8 33 -0 .0 1 -0 .0 9
10 0 .0 7 0 .9 5 22 0 .0 5 0 .6 3 34 0 .1 8 2 .3 1
11 0 .1 7 2 .1 9 23 0 .1 9 2 .5 0 35 0 .0 7 0 .9 5
12 -0 .3 2 -4 .0 8 24 0 .0 1 0 .0 9 36 -0 .0 9 -1 .1 8
1 هW'!
ذج ا1!'
ن ا132 B 78
ا7CاA
اw. وا
;ا7CاA
اw.هة
وال ا
;اK!
& ا
!ط ا
أيSARIMA(1,1,1)(0,1,1)12
166
(1 − φ B ) (1 − B12 ) (1 − B ) zt = (1 − θ B ) (1 − ΘB12 ) at
:7
;
آ4 ;!
ذج ا1!'
ا اA( ه6
MTB > ARIMA 1 1 1 0 1 1 12 'z(t)' 'RESI2';
SUBC> NoConstant;
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot.
ARIMA Model
ARIMA model for z(t)
Estimates at each iteration
Iteration SSE Parameters
0 307.653 0.100 0.100 0.100
1 281.217 0.100 0.100 0.250
2 262.275 0.226 0.231 0.400
3 262.027 0.376 0.381 0.401
4 261.770 0.526 0.531 0.401
5 261.426 0.675 0.681 0.402
6 260.905 0.824 0.831 0.403
7 260.036 0.970 0.981 0.405
8 227.926 0.835 0.980 0.536
9 221.838 0.748 0.980 0.576
10 221.665 0.738 0.980 0.586
11 221.637 0.738 0.980 0.589
12 221.610 0.737 0.980 0.589
13 221.585 0.737 0.980 0.590
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.7374 0.0620 11.89
MA 1 0.9796 0.0017 582.86
SMA 12 0.5898 0.0736 8.01
167
Differencing: 1 regular, 1 seasonal of order 12
Number of observations: Original series 178, after
differencing 165
Residuals: SS = 214.393 (backforecasts excluded)
MS = 1.323 DF = 162
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36
48
Chi-Square 15.7(DF= 9) 30.9(DF=21) 61.6(DF=33)
67.1(DF=45)
1 ه4 ;!
@ اAO
;حJ!
ذج ا1!'
أي أن ا
(1 − 0.74 B ) (1 − B12 ) (1 − B ) zt = (1 − 0.98B ) (1 − 0.59 B12 ) at , at ∼ N ( 0,1.323)
ان5X
φ = 0.74, s.e. (φ ) = 0.062, t = 11.89
θ = 0.96, s.e. (θ ) = 0.0017, t = 582.86
Θ = 0.59, s.e. ( Θ ) = 0.074, t = 8.01
.421'"!
ا4
)
"!
أي ان ا
:I> ا ا/a
\0 إ?ر ا
MTB > ZTest 0.0 1.15 'RESI1';
SUBC> Alternative 0.
Z-Test
Test of mu = 0.0000 vs mu not = 0.0000
The assumed sigma = 1.15
Variable N Mean StDev SE Mean Z P
RESI1 165 -0.0144 1.1433 0.0895 -0.16 0.87
اb+ 7Bا1(
اw1;& \ أنN X اي0.05 & ( اآ7 وهP-value=0.87 ان ا
ـ5X
Iإ?ر ا ا ا
168
MTB > Runs 0 'RESI1'.
Runs Test
RESI1
K = 0.0000
The observed number of runs = 67
The expected number of runs = 82.4061
73 Observations above K 92 below
The test is significant at 0.0149
اءL;ج إ
إ%2 اA وه7Bا1(
ا48ا1K 4PN \N ' اي ا0.05 ' ي1'"&z ;(ر9ا
:7
;
اw1
اSign Test رة9 إ;(ر اH]& ة & إ;(ر ا
ي1B ] أآ, إ;(ر
MTB > STest 0.0 'RESI1';
SUBC> Alternative 0.
Sign Test for Median
Sign test of median = 0.00000 versus not = 0.00000
N N* Below Equal Above P Median
RESI1 165 13 92 0 73 0.1611 -0.08139
w1
اWC
رات ا9 ن1 آ13
و
;{آ ي إ;(ر و0.1611 ' ي1'"& ;(ر9وا
7
;
ا
MTB > WTest 0.0 'RESI1';
SUBC> Alternative 0.
Wilcoxon Signed Rank Test
Test of median = 0.000000 versus median not = 0.000000
Number N for Wilcoxon Estimated
N Missing Test Statistic P Median
RESI1 165 13 165 6321.0 0.392 -0.05940
0.392 ' ي1'"& `2;(ر ا9وا
:Iل ا ا, 0إ?ر إ
169
78
ا7CاA
اw. وا
;ا7CاA
اw.دوال ا
;ا
A C F o f R e s id u a ls f o r z ( t)
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e a u t o c o r r e l a t io n s )
1 .0
0 .8
0 .6
Autocorrelation
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
3 6 9 12 15 18 21 24 27 30 33 36 39
Lag
P A C F o f R e s id u a ls f o r z ( t)
( w it h 9 5 % c o n f id e n c e l im it s f o r t h e p a r t ia l a u t o c o r r e l a t io n s )
1 .0
0 .8
Partial Autocorrelation
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
3 6 9 12 15 18 21 24 27 30 33 36 39
Lag
.4J; & 7ON 4"(= S وإذا آ46. &;اz O ا
(`ء أي ا4 `
ا !ط ا76"C O ا5
:I ا اi إ?ر
Histogram of the Residuals
(response is z(t))
50
40
Frequency
30
20
10
-6 -5 -4 -3 -2 -1 0 1 2 3 4
Residual
170
Normal Probability Plot of the Residuals
(response is z(t))
Residual
0
-1
-2
-3
-4
-5
-3 -2 -1 0 1 2 3
Normal Score
K-S Test for Residuals
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-5 -4 -3 -2 -1 0 1 2 3
RESI1
Average: -0.0144171 Kolmogorov-Smirnov Normality Test
StDev: 1.14327 D+: 0.117 D-: 0.140 D : 0.140
N: 165 Approximate P-Value < 0.01
ايα = 0.05 ' ي1'"& ;(ر9 اذا ا0.01 & HB أK-S ;(ر9 P-value ان ا
ـ5X
.7Bا1(
ا4"(= 4PN \N
X
: ام ا ذج:0c$ 4ا
:('C ;اتN 95% V& 4(J; & 4!B 36 :(';
. م1J'
Forecasts from period 178
95 Percent Limits
Period Forecast Lower Upper Actual
179 57.7885 55.5332 60.0437
180 55.8516 53.0220 58.6812
181 54.7429 51.6264 57.8594
182 54.1820 50.9063 57.4578
171
183 54.6298 51.2601 57.9994
184 54.9152 51.4875 58.3430
185 55.6640 52.1986 59.1294
186 59.3778 55.8869 62.8688
187 68.6924 65.1833 72.2015
188 74.0698 70.5472 77.5925
189 73.9650 70.4317 77.4983
190 63.5866 60.0447 67.1286
191 58.9095 55.1843 62.6347
192 56.7768 52.9407 60.6128
193 55.5237 51.6169 59.4305
194 54.8564 50.9022 58.8105
195 55.2256 51.2380 59.2131
196 55.4531 51.4409 59.4653
197 56.1592 52.1278 60.1905
198 59.8416 55.7947 63.8885
199 69.1329 65.0729 73.1929
200 74.4932 70.4217 78.5647
201 74.3758 70.2940 78.4575
202 63.9881 59.8968 68.0793
203 59.3041 55.0418 63.5664
204 57.1663 52.7962 61.5364
205 55.9095 51.4676 60.3514
206 55.2394 50.7472 59.7315
207 55.6066 51.0774 60.1357
208 55.8326 51.2749 60.3903
209 56.5376 51.9568 61.1184
210 60.2191 55.6189 64.8194
211 69.5099 64.8927 74.1270
212 74.8697 70.2374 79.5021
213 74.7520 70.1057 79.3983
214 64.3640 59.7046 69.0235
و
! )
. Oا
;
:7
172
8 0
7 0
Forecast
6 0
5 0
0 1 0 2 0 3 0 4 0
T im e
:(';
;ات اNات و:(';
اV& O&3. 4 ;!
7
;
ا
) ا
80
70
Forecast
60
50
0 100 200
T im e
173
ﺍﻟﻔﺼﻞ ﺍﻟﺴﺎﺑﻊ
ا9 ار ا/7jا-
ك/\ ا0 ذج ا7 ;0 ا$ 4 اN رIور
Forecasting By ARMA Models
4'& ز4 ;& H3 7 4 & 48ا1K هةI
4
;
هات اK!
ا
12.0 20.5 21.0 15.5 15.3 23.5 24.5 21.3
23.5 28.0 24.0 15.5 17.3 25.3 25.0 36.5
36.5 29.6 30.5 28.0 26.0 21.5 19.7 19.0
16.0 20.7 26.5 30.6 32.3 29.5 28.3 31.3
32.2 26.4 23.4 16.4
:7
;
اH3K
اO
و
MTB > TSPlot C1;
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
30
C1
20
10
Index 10 20 30
78
ا7CاA
اw. وا
;ا7CاA
اw. دوال ا
;اL1
Xاو
174
MTB > %ACF C1.
Autocorrelation Function for C1
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9
Lag Corr T LBQ Lag Corr T LBQ
1 0.63 3.79 15.62 8 -0.20 -0.78 31.42
2 0.30 1.35 19.27 9 -0.12 -0.44 32.10
3 0.14 0.59 20.07
4 -0.05 -0.20 20.17
5 -0.24 -1.01 22.71
6 -0.30 -1.22 26.71
7 -0.24 -0.94 29.40
MTB > %PACF C1.
Executing from file: E:\MTBWIN\MACROS\PACF.MAC
Partial Autocorrelation Function for C1
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9
Lag PAC T Lag PAC T
1 0.63 3.79 8 -0.07 -0.39
2 -0.16 -0.98 9 0.05 0.29
3 0.04 0.22
4 -0.20 -1.20
5 -0.18 -1.09
ARMA (1,1) ذج1! & ن13C B هاتK!
& ا !ط ا
ا
; ان ا5
6 -0.04 -0.23
7 0.02 0.12
;حJ!
ذج ا1!'
( ا6
MTB > Name c17 = 'RESI1'
MTB > ARIMA 1 0 1 C1 'RESI1';
SUBC> Constant;
SUBC> Forecast 5 c14 c15 c16;
SUBC> GACF;
SUBC> GPACF;
SUBC> GNormalplot.
ARIMA Model
175
ARIMA model for C1
Estimates at each iteration
Iteration SSE Parameters
0 1337.71 0.100 0.100 21.918
1 936.95 0.250 -0.049 18.193
2 849.78 0.211 -0.199 19.106
3 751.53 0.215 -0.349 18.941
4 658.66 0.266 -0.499 17.594
5 592.30 0.372 -0.649 14.890
6 580.80 0.433 -0.699 13.314
7 579.30 0.455 -0.714 12.698
8 579.11 0.464 -0.719 12.470
9 579.08 0.467 -0.721 12.386
10 579.08 0.468 -0.722 12.356
11 579.08 0.468 -0.722 12.345
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.4684 0.1755 2.67
MA 1 -0.7221 0.1380 -5.23
Constant 12.345 1.154 10.70
Mean 23.221 2.170
Number of observations: 36
Residuals: SS = 523.365 (backforecasts excluded)
MS = 15.860 DF = 33
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36 48
Chi-Square 7.2(DF=10) 15.9(DF=22) * (DF= *) * (DF=
*)
Forecasts from period 36
95 Percent Limits
Period Forecast Lower Upper
Actual
37 14.7649 6.9578 22.5720
176
38 19.2606 7.1228 31.3985
39 21.3663 8.4715 34.2610
40 22.3524 9.2975 35.4074
41 22.8143 9.7245 35.9041
:1;ح هJ!
ذج ا1!'
ا
zt = 12.345 + 0.4684 zt −1 + at − 0.7221at −1 , at ∼ WN ( 0,15.86 ) ∀t
Q5
( )
φˆ1 = 0.4684 se φˆ1 = 0.1755 t = 2.67
θˆ = −0.7221 se (θˆ ) = 0.1380 t = −5.23
1 1
δˆ = 12.345 se (δˆ ) = 1.154 t = 10.70
σˆ 2 = 15.86 df = 33
4Pb
!] اN α = 0.05 ' 421'"& راتJ!
اV!L ان5 و
H 0 : φ1 = 0
H1 : φ1 ≠ 0
φˆ1 0.4684
أيα = 0.05 ' 421'"& 7 وهt = = = 2.6689 48 59.
[;(ه
( )
se φˆ1 0.1755
ىXرات اJ
اV!
H]!
. وφ1 = 0 \ انN
'
ا
I> ا ا/a 7]
:7Bا1(
اw1;& إ;(ر
1;(ر ه9ا
H 0 : µ a = 0, H1 : µa ≠ 0
MTB > TTest 0.0 'RESI1';
SUBC> Alternative 0.
T-Test of the Mean
Test of mu = 0.000 vs mu not = 0.000
Variable N Mean StDev SE Mean T P
177
RESI1 36 0.344 3.851 0.642 0.54 0.60
أي انα = 0.05 & ( اآ7 وه0.6 7 هO
P-Value وا
ـt = 0.54 ان5X
b
وي ا2 7Bا1(
اw1;& إ;(ر3!2 ي أي1'"& z ;(ر9ا
:Iإ?ر ا ا ا
Runs Test إ;(ر ا
يf
A
و
;[م
MTB > Runs 'RESI1'.
Runs Test
RESI1
K = 0.3443
The observed number of runs = 21
The expected number of runs = 19.0000
18 Observations above K 18 below
The test is significant at 0.4989
Cannot reject at alpha = 0.05
α = 0.05 ' 7Bا1(
ا48ا1K \N'' ر3!2X
:I\ ا ا$إ?ر
ا
7CاA
اw. إ;(ر ا
;اf
A
و
;[م
ACF of Residuals for C1
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9
Lag
PACF of Residuals for C1
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9
Lag
178
7K!;C ا
!طOIC O
أي ا7Bا1(
4b;[!
) اJ
ا. 4L & أي درw.اC أيL12X *
ا5
`ء. 4 P 4 ;& O
1 آV&
Normal Probability Plot
ـ. 7Bا1(
ا4"(= (;[ واا
Normal Probability Plot of the Residuals
(response is C1)
10
Residual
-10
-2 -1 0 1 2
Normal Score
( 1 ) ل1(J& 1وه
('& ;حJ!
ذج ا1!'
إ;(ر ا3!2 إذا
O
:('C 95% ;اتN و4(J; & )B 4 ![
ات:(';
7
;
ا
) ا
MTB > TSPlot C14 C15 C16;
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect;
SUBC> overlay.
Time Series Plot for C1
(with forecasts and their 95% confidence limits)
36
26
C1
16
6
5 10 15 20 25 30 35
Time
179
:
?z ل-
4'& ز4 ;& H3 7 4 & 48ا1K هةI
4
;
هات اK!
ا
10.38 11.86 10.97 10.80 9.79 10.39 10.42 10.82
11.40 11.32 11.44 11.68 11.17 10.53 10.01 9.91
9.14 9.16 9.55 9.67 8.44 8.24 9.10 9.09
9.35 8.82 9.32 9.01 9.00 9.80 9.83 9.72
9.89 10.01 9.37 8.69 8.19 8.67 9.55 8.92
8.09 9.37 10.13 10.14 9.51 9.24 8.66 8.86
8.05 7.79 6.75 6.75 7.82 8.64 10.58 9.48
7.38 6.90 6.94 6.24 6.84 6.85 6.90 7.79
8.18 7.51 7.23 8.42 9.61 9.05 9.26 9.22
9.38 9.10 7.95 8.12 9.75 10.85 10.41 9.96
9.61 8.76 8.18 7.21 7.13 9.10 8.25 7.91
6.89 5.96 6.80 7.68 8.38 8.52 9.74 9.31
9.89 9.96
MTB > TSPlot C10;
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
12
11
10
10
9
C
Index 10 20 30 40 50 60 70 80 90
78
ا7CاA
اw. وا
;ا7CاA
اw. دوال ا
;اo%b
180
Autocorrelation Function for C10
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 12 22
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 0.83 8.24 69.92 8 0.26 1.26 178.83 15 0.05 0.21 191.63 22 0.10 0.47 194.02
2 0.61 3.91 107.90 9 0.26 1.21 186.14 16 0.04 0.16 191.78 23 0.18 0.81 198.12
3 0.46 2.56 129.56 10 0.18 0.84 189.86 17 0.00 0.02 191.78 24 0.20 0.89 203.24
4 0.37 1.95 143.87 11 0.09 0.43 190.87 18 -0.03 -0.15 191.91
5 0.33 1.65 155.04 12 0.04 0.20 191.09 19 -0.05 -0.24 192.26
6 0.28 1.40 163.68 13 0.03 0.13 191.19 20 -0.05 -0.24 192.60
7 0.26 1.28 171.23 14 0.04 0.19 191.39 21 0.01 0.07 192.63
Partial Autocorrelation Function for C10
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 12 22
Lag PAC T Lag PAC T Lag PAC T Lag PAC T
1 0.83 8.24 8 0.05 0.45 15 -0.01 -0.15 22 0.05 0.51
2 -0.27 -2.64 9 0.00 0.03 16 -0.03 -0.25 23 0.06 0.59
3 0.13 1.29 10 -0.20 -1.98 17 -0.07 -0.73 24 -0.07 -0.65
4 0.03 0.34 11 0.02 0.19 18 -0.03 -0.26
5 0.06 0.61 12 0.01 0.09 19 0.06 0.60
6 -0.02 -0.21 13 0.01 0.12 20 0.02 0.20
7 0.09 0.91 14 0.03 0.34 21 0.21 2.03
AR ( 2 ) 4
]
ا4L & ا
ر7Cار ذا%
ذج إ1!
;حJC ان ا
!ط5
;حJ!
ذج ا1!'
( ا6
MTB > Name c17 = 'RESI1'
MTB > ARIMA 2 0 0 C10 'RESI1';
SUBC> Constant;
SUBC> Forecast 5 c14 c15 c16;
SUBC> GSeries;
SUBC> GACF;
SUBC> GPACF;
SUBC> GNormalplot.
ARIMA Model
181
ARIMA model for C10
Estimates at each iteration
Iteration SSE Parameters
0 126.398 0.100 0.100 7.283
1 103.515 0.250 0.043 6.434
2 84.535 0.400 -0.014 5.586
3 69.407 0.550 -0.071 4.738
4 58.132 0.700 -0.128 3.887
5 50.724 0.850 -0.184 3.030
6 47.212 1.000 -0.239 2.163
7 46.918 1.053 -0.256 1.838
8 46.916 1.054 -0.255 1.816
9 46.916 1.054 -0.255 1.814
10 46.916 1.054 -0.255 1.814
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
AR 1 1.0542 0.0992 10.63
AR 2 -0.2547 0.0993 -2.56
Constant 1.81360 0.07092 25.57
Mean 9.0480 0.3538
Number of observations: 98
Residuals: SS = 46.7518 (backforecasts excluded)
MS = 0.4921 DF = 95
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36 48
Chi-Square 7.2(DF=10) 13.7(DF=22) 21.3(DF=34)
28.8(DF=46)
Forecasts from period 98
95 Percent Limits
Period Forecast Lower Upper
Actual
182
99 9.7950 8.4198 11.1703
100 9.6033 7.6050 11.6016
101 9.4432 7.1234 11.7629
102 9.3232 6.8446 11.8018
103 9.2375 6.6825 11.7925
1;ح هJ!
ذج ا1!'
ا
zt = 1.8136 + 1.0542 zt −1 − 0.2547 zt −2 + at , at ∼ WN ( 0,0.4921) ∀t
7 هO
t )B و42 ا
!"رOCNا%
رات ا
!"
) وإJ&
( )
φˆ1 = 1.0542 se φˆ1 = 0.0992 t = 10.63
φˆ = −0.2547 se (φˆ ) = 0.0993 t = −2.56
2 2
δˆ = 1.8136 se (δˆ ) = 0.07092 t = 25.57
σˆ 2 = 0.4921 df = 95
α = 0.05 ' 421'"& راتJ!
اV!L ان5
7Bا1(
ا7 ي إ;(رات:(';
W'& * ا7 ذج1!'
ا اAO. H(J 73
MTB > TTest 0.0 'RESI1';
SUBC> Alternative 0.
T-Test of the Mean
Test of mu = 0.0000 vs mu not = 0.0000
Variable N Mean StDev SE Mean T P
RESI1 98 -0.0082 0.6942 0.0701 -0.12 0.91
42b
ا4Pb
;!ل ان ا5 إ7 هP-Value ا
ـ4I5 & ) ي1'"& z ا ان ا;(رL ^Pوا
( 4%%+
7Bا1(
ا48ا1K (;[
MTB > Runs 'RESI1'.
Runs Test
RESI1
K = -0.0082
The observed number of runs = 47
The expected number of runs = 49.9184
183
47 Observations above K 51 below
The test is significant at 0.5529
Cannot reject at alpha = 0.05
7Bا1(
ا48ا1K 4PN \N'' ر3!2X أي
78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;اo%b
ACF of Residuals for C10
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 4 6 8 10 12 14 16 18 20 22 24
Lag
PACF of Residuals for C10
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 4 6 8 10 12
Lag
14 16 18 20
ا
(`ء4 `
ا ا
!ط اL ^Pوا
22 24
7Bا1(
ا4"(= o%N 7J(2
184
Normal Probability Plot of the Residuals
(response is C10)
Residual
0
-1
-2
-3 -2 -1 0 1 2 3
Normal Score
( at ∼ IIDN ( 0,0.4921) ) أي7"(= V2ز1C O
7Bا1(
ل ان ا1J انV6; و
O
:('C 95% ;اتN و4(J; & )B 4 ![
ات:(';
7
;
ا
) ا
MTB > TSPlot C14 C15 C16;
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect;
SUBC> overlay.
Time Series Plot for C10
(with forecasts and their 95% confidence limits)
12
11
10
C10
10 20 30 40 50 60 70 80 90
Time
_8;'
ا. رنB وAR (1) ذج1!
4J.
هات اK!
ا7 (= : 2!C
185
186
ﺍﻟﻔﺼﻞ ﺍﻟﺜﺎﻣﻦ
N0 و
إ?ر ا ذج اI ا ا./ ' ل-
: Example on Residual Analysis and Model Selection Criteria
ة6"& هاتK& !N 4J(6!
) اJ
اoB هةK!
) اJ
اO ا7 7Bا1(
اJ. 'N J
:7Bا1(
اW;3C وzˆ1 , zˆ2 ,..., zˆn 4J(6& )B '2
_;'2 (6& ذج1!
وz1 , z2 ,..., zn
ei = zi − zˆi , i = 1,2,..., n
J%C انW 2 اAO
وaˆi = ei , i = 1,2,..., n ذج أي1!'
ا7N ء6رات اJ& 7 ه7Bا1(
وا
:O'& 7;
ذج وا1!'
ا اA ه7N ء6 ا7 4Pوb!
وط اK
ا
b
وي ا2 ء6 اw1;& -1
ء6;ض ان اb
آ] & ا
'!ذج7N ) و4J; & أو46. &;اz و48ا1K ء6 ا-2
( at ∼ IIDN ( 0,σ 2 ) أيσ 2 2(Cي وb+ w1;!. .6;& وHJ; & 7"(= V2ز1C O
@A هJ%C S ! إذا آN
'ي7Bا1(
ا7 ;(رات9 & ا41! & 1 وه%C ' يFN اAO
W N ;(رات9@ اA ه5 اHKN أ& إذاX1(J& (6!
ذج ا1!'
";( ا4
%
@ اA ه7Nوط وK
ا
, ذج1! ;احB وإI'
' إدة ا
w1;!
إ;(ر ا:Xأو
H 0 : E ( at ) = 0
H 1 : E ( at ) ≠ 0
e
'"N 7B 7"(= V2ز1C O
7;
واu = 48 59 ا4N و
;[م2A. إ;(ر1وه
se ( e )
إ;(ر ان7 اA ) هu < 1.96 S
إذا آE ( at ) = 0
";( انα = 0.05 421'"& ى1; &
( O
ر7;
ا4'&
!; ت اJ%;& !8ا داAة وه5 و30 & ( اآ4'"
) ا5
:ل-
Metals Y;!
ا7 4]
]
ا4Lك & ا
ر%;& w1;& (6C &]ل7
د ا1"
ف1
MTB > RETR 'E:\Mtbwin\DATA\EMPLOY.MTW'.
Retrieving worksheet from file: E:\Mtbwin\DATA\EMPLOY.MTW
Worksheet was saved on 6/ 5/1996
MTB > TSPlot 'Metals';
SUBC> Index;
187
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
e
M ls
ta 50
45
40
Index 10 20 30 40 50 60
MTB > Name c4 = 'AVER1' c5 = 'FITS1' c6 = 'RESI1'
MTB > %MA 'Metals' 3;
SUBC> Averages 'AVER1';
SUBC> Fits 'FITS1';
SUBC> Residuals 'RESI1'.
Executing from file: E:\MTBWIN\MACROS\MA.MAC
Moving average
Data Metals
Length 60.0000
NMissing 0
Moving Average
Length: 3
Accuracy Measures
MAPE: 1.55036
MAD: 0.70292
MSD: 0.76433
188
Moving Average
Actual
Predicted
50
Actual
Predicted
Metals
45
Moving Average
Length: 3
MAPE: 1.55036
MAD: 0.70292
40 MSD: 0.76433
0 10 20 30 40 50 60
Time
~&[
د ا1!"
ا7N 4J(6!
) اJ
د ا
دس وا1!"
ا7N 7Bا1(
ا
'
ا5X
MTB > print c3 c6 c5
Data Display
Row Metals RESI1 FITS1
1 44.2 * *
2 44.3 * *
3 44.4 * *
4 43.4 -0.90000 44.3000
5 42.8 -1.23333 44.0333
6 44.3 0.76667 43.5333
7 44.4 0.90000 43.5000
8 44.8 0.96667 43.8333
9 44.4 -0.10000 44.5000
10 43.1 -1.43333 44.5333
11 42.6 -1.50000 44.1000
12 42.4 -0.96667 43.3667
13 42.2 -0.50000 42.7000
14 41.8 -0.60000 42.4000
15 40.1 -2.03333 42.1333
189
16 42.0 0.63333 41.3667
17 42.4 1.10000 41.3000
18 43.1 1.60000 41.5000
19 42.4 -0.10000 42.5000
20 43.1 0.46667 42.6333
21 43.2 0.33333 42.8667
22 42.8 -0.10000 42.9000
23 43.0 -0.03333 43.0333
24 42.8 -0.20000 43.0000
25 42.5 -0.36667 42.8667
26 42.6 -0.16667 42.7667
27 42.3 -0.33333 42.6333
28 42.9 0.43333 42.4667
29 43.6 1.00000 42.6000
30 44.7 1.76667 42.9333
31 44.5 0.76667 43.7333
32 45.0 0.73333 44.2667
33 44.8 0.06667 44.7333
34 44.9 0.13333 44.7667
35 45.2 0.30000 44.9000
36 45.2 0.23333 44.9667
37 45.0 -0.10000 45.1000
38 45.5 0.36667 45.1333
39 46.2 0.96667 45.2333
40 46.8 1.23333 45.5667
41 47.5 1.33333 46.1667
42 48.3 1.46667 46.8333
43 48.3 0.76667 47.5333
44 49.1 1.06667 48.0333
45 48.9 0.33333 48.5667
46 49.4 0.63333 48.7667
47 50.0 0.86667 49.1333
48 50.0 0.56667 49.4333
49 49.6 -0.20000 49.8000
190
50 49.9 0.03333 49.8667
51 49.6 -0.23333 49.8333
52 50.7 1.00000 49.7000
53 50.7 0.63333 50.0667
54 50.9 0.56667 50.3333
55 50.5 -0.26667 50.7667
56 51.2 0.50000 50.7000
57 50.7 -0.16667 50.8667
58 50.3 -0.50000 50.8000
59 49.2 -1.53333 50.7333
60 48.1 -1.96667 50.0667
7Bا1(
اw1;& (;[
نsا
MTB > TTest 0.0 'RESI1';
SUBC> Alternative 0.
T-Test of the Mean
Test of mu = 0.000 vs mu not = 0.000
Variable N Mean StDev SE Mean T P
RESI1 57 0.158 0.868 0.115 1.37 0.17
z ( 2(;
اف ا
!"ري ) او ا%
9ن ا132 &'
&_ م. ;[م2 Minitab 7N :4I5 &
أي1.96 & HB ا7 وهT=1.37 7 ه48 59 ا4!B ان5X . Ttest * 62&"وف و
42b
ا4Pb
\ اN
X
b
ل ا15 وw1;!
ل ا15 Runs test إ;(ر ا
ي46ا1. 7Bا1(
ا48ا1K (;[
:
U
: ا
!]لV.C
MTB > Runs 'RESI1'.
Runs Test
191
RESI1
K = 0.1579
The observed number of runs = 17
The expected number of runs = 29.4211
30 Observations above K 27 below
The test is significant at 0.0009
MTB > Runs 0 'RESI1'.
Runs Test
RESI1
K = 0.0000
The observed number of runs = 17
The expected number of runs = 28.7895
33 Observations above K 24 below
The test is significant at 0.0013
7Bا1(
ا48ا1K \N
X ;
%
آ; ا7N *
ا5
Autocorrelation test 7CاA
اw. إ;(ر ا
;ا46ا1. 7Bا1(
ل اJ; أو إw.اC (;[
:]
U
: ا
!]لV.C
MTB > %ACF 'RESI1'.
Executing from file: E:\MTBWIN\MACROS\ACF.MAC
192
Autocorrelation Function for RESI1
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
4 9 14
Lag Corr T LBQ Lag Corr T LBQ
1 0.56 4.24 18.93 8 -0.04 -0.22 24.17
2 0.24 1.39 22.31 9 0.12 0.68 25.21
3 -0.01 -0.06 22.32 10 0.29 1.64 31.36
4 0.04 0.21 22.40 11 0.24 1.30 35.67
5 0.02 0.11 22.42 12 0.20 1.05 38.71
6 -0.07 -0.38 22.72 13 0.05 0.25 38.90
7 -0.14 -0.81 24.07 14 0.03 0.17 38.99
أيρ1 = 0 \ انN ' أي ا4.24 ويC اولe[;
' ا7CاA
اw.
;اT ان ا
ـ5X
;(ر9 ا7N
وX ا4L & ا
ر7Bا1(
ا. w.اC L12
H 0 : ρ1 = 0
H1 : ρ1 ± 0
r1
= 4.24 7 ه48 59 اQ5
se ( r1 )
"(= 4ز1& 7Bا1(
اS & إذا آ7N (;[ :".را
: ا
!]لV.C
MTB > %NormPlot 'RESI1';
SUBC> Kstest.
Executing from file: E:\MTBWIN\MACROS\NormPlot.MAC
Normal Probability Plot
.999
.99
.95
ility
.80
bab
.50
Pro
.20
.05
.01
.001
-2 -1 0 1
RESI1
Average: 0.157895 Kolmogorov-Smirnov Normality Test
StDev: 0.867525 D+: 0.054 D-: 0.084 D : 0.084
N: 57 Approximate P-Value > 0.15
193
V2ز1;
ا4PN \N
X أي0.05 & ( اآ7 وه0.15 ويC 4 C'
اP-Value ا
ـ5X
( &ي2 يA
واQ-Q Plot ا
ـ1 ه4"(6
, ` إ;(ر2 ه'ك اα = 0.05 ' 7"(6
ا
"& V2ز1C V& & هاتK& .6C
MTB > %Qqplot 'RESI1';
SUBC> Table;
SUBC> Conf 95;
SUBC> Ci.
Executing from file: E:\MTBWIN\MACROS\Qqplot.MAC
Distribution Function Analysis
Normal Dist. Parameter Estimates
Data : RESI1
Mean: 0.157895
StDev: 0.867525
Normal Probability Plot for RESI1
99
Mean: 0.157895
StDev: 0.867525
95
90
80
70
Percent
60
50
40
30
20
10
-2 -1 0 1 2
Data
7"(= V2ز1C O
7Bا1(
\ ان اN
X '
FN ;
%
آ; ا7N ان5X
194
& 4I5اة(2 :و ان ا
(1ا )I"& J%C 7Bا
Kوط !Nي ا
;ا w.ا
Aي . L12ا
)J
ا
!;;
4وهAا 1L \N
'" 2دة ا
; 4J26
(6ا
!; w1ا
!;%ك & ا
ر 4Lا
]
]Q5 4
ادي ا
1. 7ا;& 7Bا.46.
195
ﺍﻟﻔﺼﻞ ﺍﻟﺘﺎﺳﻊ
:Decomposition Method )| ا ا
آتY او./
213;
O`". V& ة%;& اءL & ة &آ(ت أو ا4
13& O
ا4'&
ا4 ;!
إ
اI'2
،4 ;!
@ اAه
O;LA!
3!2 *
أ4. ;
. L وJ
. z1 , z2 ,..., zn @هK!
ا4'&
ا4 ;!
' ا2
;ض انb'
H3K
ا
zt = Tt + St + Ct + Et , t = 1, 2,..., n
يA
@ ا
"م اC9ج اA!'C 7;
ا7
اف وه9 ا4( &آTt هة وK!
ا4'&
ا4 ;!
اZt Q5
1( وهL )إذا و7!1!
اU{;
ج اA!'C و4!1& 4( &آSt و4 ;!
' ف ا
* اC أو7%'C
4( &آCt و421'
وا42OK
اH]& 4!1!
ات اU{;
ا4 ;
4 ;!
ث%2 يA
اY;
ا
Et و4!1& z 421= 4'&;ات زN ". ر3;2 ةC أو إ%'& جA!'C ت( وL )إذا و42دور
O;LA!
3!2X 7;
وا4 ;!
اU:C 7;
ى اX اH&ا1"
اV!L H!KC{ و6[
ا4(&آ
Additive 7NP9ذج ا1!'
. 7! 2 .
ذج ا1!'
ا.OB اوOCهK& 3!2X 7;
!' أو اP
H]& ل اى3 ه'ك أ.ذج1!'
ا7N 7NP إH3K. HC ا
!آ(تH ن آf
وذModel
zt = Tt St + Ct + Et , t = 1, 2,..., n
zt = Tt St Ct + Et , t = 1, 2,..., n
. Multiplicative Models 4b`;
'!ذج ا. ! C 7;
وا
ن13C& درا42 ا
ور4( ن ا
!آf
وذCt 42 ا
ور4( ا
!آH!O ف1 ى1; !
ا اA ه7N
اL 421= هاتK&
;ج ا%C O ( 4216
ة أو اJ
ا
!; ت ا7N دة1L1&
.د1J"
&ي د آ( & ا
H3K
ا7
'!ذج. 7b;3 و
zt = Tt + St + Et , t = 1, 2,..., n
zt = Tt St + Et , t = 1, 2,..., n
b
:!
FORECASTING: METHODS AND APPLICATIONS آ;بI أ
141-131 صMAKRIDAKIS/ WHEELWRIGHT/ McGEE
أي42 ا
ور4(ون ا
!آ. 4b`;
وا4NP9 اH%;
=ق ا7
;
ا
!]ل ا7N ف
;"ض1
ا
'!ذج
196
zt = Tt + St + Et , t = 1, 2,..., n
zt = Tt St + Et , t = 1, 2,..., n
1960 4' & 'ة3. 12 او ;ر4'2& 7N ا
;ات2 !. 2'(
ا7 W6
ا7 ه4
;
ا
( ت ا
1975 4' ;5و
GasDemand
MONTHLY GASOLINE DEMAND ONTARIO GALLON MILLIONS 1960-1975
87695 86890 96442 98133 113615 123924 128924 134775 117357 114626
107677 108087 92188 88591 98683 99207 125485 124677 132543
140735 124008 121194 111634 111565 101007 94228 104255 106922
130621 125251 140318 146174 122318 128770 117518 115492 108497
100482 106140 118581 132371 132042 151938 150997 130931 137018
121271 123548 109894 106061 112539 125745 136251 140892 158390
148314 144148 140138 124075 136485 109895 109044 122499 124264
142296 150693 163331 165837 151731 142491 140229 140463 116963
118049 137869 127392 154166 160227 165869 173522 155828 153771
143963 143898 124046 121260 138870 129782 162312 167211 172897
189689 166496 160754 155582 145936 139625 137361 138963 155301
172026 165004 185861 190270 163903 174270 160272 165614 146182 137728
148932 156751 177998 174559 198079 189073 175702 180097 155202 174508
154277 144998 159644 168646 166273 190176 205541 193657 182617 189614
174176 184416 158167 156261 176353 175720 193939 201269 218960 209861
198688 190474 194502 190755 166286 170699 181468 174241 210802 212262
218099 229001 203200 212557 197095 193693 188992 175347 196265 203526
227443 233038 234119 255133 216478 232868 221616 209893 194784 189756
193522 212870 248565 221532 252642 255007 206826 233231 212678 217173
199024 191813 195997 208684 244113 243108 255918 244642 237579 237579
217775 227621
7'& زw6[& 7N 4 ;!
) اXأو
MTB > TSPlot 'GasDemand';
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
197
250
200
GasDemand
150
100
Index 50 100 150
X ا
ا4N '& و4!1& 4'&
ا4 ;!
ان ا5
zt = Tt + St + Et , t = 1, 2,..., n :7NP9ذج ا1!'
( ا6C :Xاو
MTB > %Decomp 'GasDemand' 12;
SUBC> Additive ;
SUBC> Forecasts 24;
SUBC> Title "Forecast of Gasoline Demand";
SUBC> Start 1.
Time Series Decomposition
Data GasDeman
Length 192.000
NMissing 0
Trend Line Equation
Yt = 96.4074 + 0.680579*t
Seasonal Indices
Period Index
1 -20.5625
2 -26.8125
3 -14.8958
4 -11.0625
198
5 9.89583
6 11.8958
7 22.7708
8 25.1875
9 5.64583
10 7.27083
11 -4.81250
12 -4.52083
Accuracy of Model
MAPE: 3.6952
MAD: 5.6622
MSD: 52.7851
Forecasts
Row Period Forecast
1 193 207.197
2 194 201.627
3 195 214.225
4 196 218.738
5 197 240.377
6 198 243.058
7 199 254.614
8 200 257.711
9 201 238.850
10 202 241.155
11 203 229.753
12 204 230.725
13 205 215.364
14 206 209.794
15 207 222.391
16 208 226.905
17 209 248.544
18 210 251.225
19 211 262.780
20 212 265.878
21 213 247.017
22 214 249.322
23 215 237.919
24 216 238.892
199
(1) H3
Forecast of Gasoline Demand
Seasonal Indices Original Data, by Seasonal Period
30
250
20
10 200
0
-10 150
-20
100
-30
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
Percent Variation, by Seasonal Period Residuals, by Seasonal Period
30
20
10
10
0
5
-10
-20
0 -30
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
(2) H3
Forecast of Gasoline Demand
Original Data Detrended Data
50
250
40
30
200 20
10
0
150 -10
-20
100 -30
-40
0 100 200 0 100 200
Seasonally Adjusted Data Seasonally Adj. and Detrended Data
250 30
20
200 10
0
150 -10
-20
100 -30
0 100 200 0 100 200
(3) H3
200
Forecast of Gasoline Demand
Actual
250 Predicted
Forecast
Actual
GasDeman
200 Predicted
Forecast
150
100
MAPE: 3.6952
MAD: 5.6622
MSD: 52.7851
0 100 200
Time
:_8;'
ا4KB'&
(2 ا & ا
رH3K
N ، Seasonal Indices 4!1!
ات ا:!
^ اP12 (1) H3
oJ
ث%2 4 و3 و2 و1 و12 و11 O ا7bN 4'
& ا4b;[!
اO ا7N W6
اU{C
^( 2 ;5 2;ا2 )U 2 OK
ا7N *
&"لHB إ
أH 2 ;5 2رC oB';2 إذW6
ا7N
( آH3K. oJ'2 )U 8 OK
ا7N 4(L1& 4!B B اH 2 ;5 2;ا2 و5 OK
ا7N (L1&
4ز1& 4+هات اK!
Box Plot ر) ا
'وق76"2 !
ا & اH3K
ا.A8".
. Out Liers 4L) ا
[رJ
واO Hهات آK!
ر اK;
وإV2ز1C ^P12 1 وهO ا
H3K
ا.(O )ا4!1!
;ات اb
ي ا1!
ا7( '
اY;
ا76"2 & ا
رHb اH3K
ا
.O ا4ز1& ء6 أو ا7Bا1(
ر) ا
'وق76"2 !2 اHbا
76"2 ا & ا
رH3K
ا،4+هات اK!
ا76"2 !
ا & اH3K
( ا2) H3
اف أي9 ا4( &آ45" إزا. هاتK!
ا
wt = zt − Tt , t = 1, 2,..., n
=St + Et , t = 1, 2,..., n
أي4!1!
ا4( ا
!آ45" إزا. 4+هات اK!
ا76"2 & ا
رHb اH3K
ا
yt = zt − St , t = 1, 2,..., n
=Tt + Et , t = 1, 2,..., n
اف9 ا7;( &آ45" إزا. 7Bا1(
أو ا Et {6[
ا4( &آ76"2 !2 اHb اH3K
ا
أي4+هات اK!
& ا4!1!
وا
et = zt − Tt − St , t = 1, 2,..., n
=Et , t = 1, 2,..., n
.(6;
ا4B~ د2J& V& 4(J; !
ا24 )J
ات:(';
ا76"2 (3) H3
201
zt = Tt St + Et , t = 1, 2,..., n :7b`;
ذج ا1!'
( ا6C :
U
MTB > %Decomp 'GasDemand' 12;
SUBC> Forecasts 24;
SUBC> Title "Forecast of Gasoline Demand";
SUBC> Start 1.
Executing from file: D:\MTBWIN\MACROS\Decomp.MAC
Macro is running ... please wait
Time Series Decomposition
Data GasDeman
Length 192.000
NMissing 0
Trend Line Equation
Yt = 96.4074 + 0.680579*t
Seasonal Indices
Period Index
1 0.860355
2 0.828555
3 0.892431
4 0.936273
5 1.06124
6 1.07274
7 1.15775
8 1.17075
9 1.03409
10 1.05059
11 0.966300
12 0.968923
202
Accuracy of Model
MAPE: 3.6338
MAD: 5.7720
MSD: 56.8996
Forecasts
Row Period Forecast
1 193 195.954
2 194 189.275
3 195 204.474
4 196 215.156
5 197 244.596
6 198 247.977
7 199 268.415
8 200 272.227
9 201 241.154
10 202 245.718
11 203 226.660
12 204 227.935
13 205 202.980
14 206 196.042
15 207 211.762
16 208 222.803
17 209 253.263
18 210 256.738
19 211 277.870
20 212 281.789
21 213 249.599
22 214 254.298
23 215 234.552
24 216 235.848
(4) H3
203
Forecast of Gasoline Demand
Seasonal Indices Original Data, by Seasonal Period
1.2
250
1.1
200
1.0
150
0.9
100
0.8
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
Percent Variation, by Seasonal Period Residuals, by Seasonal Period
14
20
12
10 10
8 0
6
-10
4
(5) H3
2 -20
0 -30
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
Forecast of Gasoline Demand
Original Data Detrended Data
250 1.3
1.2
200
1.1
150 1.0
0.9
100 0.8
0 100 200 0 100 200
Seasonally Adjusted Data Seasonally Adj. and Detrended Data
240
20
220
200 10
180 0
160
-10
140
(6) H3
120 -20
100 -30
0 100 200 0 100 200
Forecast of Gasoline Demand
Actual
280
Predicted
Forecast
Actual
GasDeman
Predicted
Forecast
180
MAPE: 3.6338
MAD: 5.7720
80 MSD: 56.8996
0 100 200
Time
:_8;'
ا4KB'&
.(3) ( و2) ( و1) ل3 ا7N ! آb;
~ اb
O
(6) ( و5) ( و4) ل3ا
204
دور7C{2 ' وه،ذج1!
H`N أن
[;ر اW N هةK!
ا4 ;!
اLذ1!
'J(= '
! ا.
:4B~ د2J& 4U U '2
،_&
(
';_ & اC 7;
( وا6;
ا4B~ د2J&
MAPE أوMean Absolute Percentage Error 6!
ا7( '
{ ا6[
اw1;& -1
4B "
. 6"2و
n
zt − zˆt
∑ t =1 zt
MAPE = × 100, zt ≠ 0
n
4B "
. 6"2 وMAD أوMean Absolute Deviation 6!
اف ا%
9 اw1;& -2
n
∑z t − zˆt
MAD = t =1
n
4B "
. 6"2 ( وMSE )أوMSD (V.!
{ ا6[
اw1;& )أوV.!
اف ا%
9 أw1;& -3
n
∑( z − zˆt )
2
t
MSD = t =1
n
]س اآJ!
ا،سJ!
ا اAO
4!B HB أ76"2 يA
ذج ا1!'
~
[;ر ا2J!
ة اA ه5;ر أF.
.'ف ه1 يA
ا1 وهMSE أوMSD 1 ه1إ;[ا& و
:7 ه4B
~ ا2J& 7NP9ذج ا1!'
MAPE: 3.6952
MAD: 5.6622
MSD: 52.7851
:7b`;
ذج ا1!'
و
MAPE: 3.6338
MAD: 5.7720
MSD: 56.8996
ذج1!'
ا اAر إ;[ام هJ
f
A
وMSD سJ!
4!B HB ا6 ا7NP9ذج ا1!'
أن ا5
.W6
ا4 ;!
4(J; !
) اJ
ا:(';
205
:Decomposition Method )| ا ا
آتY او./
i }A
إ
;ج4
;
ا
(
ت ا. إ
&آ(ت4 ;!
اf3bC أوH%C 4J2= 7
;
!]ل ا. ^P1
ف1
امL 13
. ا
!ارع5 أ7N W%
&12 168
MTB > Read "E:\Mtbwin\milk.dat" c1.
Entering data from file: E:\Mtbwin\milk.dat
168 rows read.
MTB > name c1='MilkProd'
MTB > print c1
Data Display
MilkProd
589 561 640 656 727 697 640 599 568 577
553 582 600 566 653 673 742 716 660 617
583 587 565 598 628 618 688 705 770 736
678 639 604 611 594 634 658 622 709 722
782 756 702 653 615 621 602 635 677 635
736 755 811 798 735 697 661 667 645 688
713 667 762 784 837 817 767 722 681 687
660 698 717 696 775 796 858 826 783 740
701 706 677 711 734 690 785 805 871 845
801 764 725 723 690 734 750 707 807 824
886 859 819 783 740 747 711 751 804 756
860 878 942 913 869 834 790 800 763 800
826 799 890 900 961 935 894 855 809 810
766 805 821 773 883 898 957 924 881 837
784 791 760 802 828 778 889 902 969 947
908 867 815 812 773 813 834 782 892 903
966 937 896 858 817 827 797 843
:&وا. 4J.
و
) ا
(
ت ا
MTB > TSPlot 'MilkProd';
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
206
1000
900
MilkProd
800
700
600
Index 50 100 150
:7NP9ذج ا1!'
( ا6C :Xاو
73
zt = Tt + St + Et , t = 1, 2,..., n H3K
ذج ا1!'
اW;3
z1 , z2 ,..., zn هاتK!
:4'3!& ;J2= ف ;"ض1 4J.
ا
!آ(ت ا7
ا4'&
ا4 ;!
@ اA هf3b;. م1J
:7
وX ا4J26
ا
: أيTt افX ا4( &آ2J;
&
ا7 هاتK!
w . 76 ار% ( إ6 -1
Tˆt ≡ zˆt = a + bt , t = 1, 2,...,168
:أي
MTB > set c2
DATA> 1:168
DATA> end
MTB > name c1='MilkProd' c2='Time' c3='Trend'
c5='Detrend' c6='Index' c8='Fitted' c9='Resid'
MTB > regr c1 1 c2;
SUBC> fits c3.
Regression Analysis
The regression equation is
MilkProd = 612 + 1.69 Time
Predictor Coef StDev T P
Constant 611.682 9.414 64.97 0.000
Time 1.69262 0.09663 17.52 0.000
207
S = 60.74 R-Sq = 64.9% R-Sq(adj) = 64.7%
Analysis of Variance
Source DF SS MS F
P
Regression 1 1132003 1132003 306.83
0.000
Error 166 612439 3689
Total 167 1744443
1
اف ه9 اH3و
900
800
Trend
700
600
Index 50 100 150
4
&ا4 ;!
. 7! 2& 7 H %'N 4+Xهات اK!
اف & ا9 ا4(ح &آ6
-2
zt − zˆt = zt − Tˆt , t = 1,2,...,168 أيDetrended Series
اف9ا
MTB > let c5=c1-c3
:7
;
اH3K
اO
و
100
Detrend
-100
Index 50 100 150
208
zt − Tˆt = St + Et , t = 1, 2,...,168 نwJN 4!1& S%(+ اO
ا5
:7
;
آSeasonal Indices 4!1!
ات ا:!
اL1
4!1!
ا4( ا
!آ2J;
-3
OK
7!1!
ا:!
اI1 Q5 I s , s = 1,2,...,12 &
. 4!1!
ات ا:!
&'
d t = zt − Tˆt , t = 1, 2,...,168 ـ. &'
ا وA3 وه7 ]
اOK
7!1!
ا:!
اI 2 اول و
:7
;
ات آ:!
@ اAر هJC
1
I1 = ( d1 + d13 + d 25 + ⋯ + d157 )
14
1
I 2 = ( d 2 + d14 + d 26 + ⋯ + d158 )
14
⋮
1
I12 = ( d12 + d 24 + d 36 + ⋯ + d168 )
14
:4
;
;[ام اوا& اF. f
;) ذ2و
MTB > set c4
DATA> 14(1:12)
DATA> end
MTB > stat c5;
SUBC> by c4;
SUBC> mean c6.
MTB > Stack 'Index' 'Index' 'Index' 'Index' 'Index'
'Index' 'Index' &
CONT> 'Index' 'Index' 'Index' 'Index' 'Index' 'Index'
'Index' c7.
MTB > let c8=c3+c7
MTB > let c9=c1-c8
MTB > set c10
DATA> 1:12
DATA> end
MTB > print c10 c6
Data Display
209
Row Season Index
1 1 -18.328
2 2 -57.806
3 3 34.716
4 4 49.595
5 5 110.616
6 6 82.281
7 7 32.517
8 8 -9.747
9 9 -52.297
10 10 -48.775
11 11 -79.754
12 12 -43.018
-4ا
;'(:ات
1Cآ
;
:7
z168 ( ℓ ) = 612 + 1.69 ( ℓ + 168 ) + I ( ℓ mod 12 ) , ℓ = 1, 2,...
]!Nا
;'( ' :ا
1م 169ه1
z168 (1) = 612 + 1.69 (169 ) + I1
=897.61 + ( −18.328 ) = 879.282
ا
4J26ا
]
:4
وه 7ا
;: Minitab _&
. O&[; 2 7
-1آ
4J26اXو
(6إ
%ار K!
w . 76هات 7ا
&
;& 2Jآ( 4ا
Xاف
~b
H %'N Ttا
'; 4آ! 7Nا
4J26اXو
)(1
-2ا `2ه'
6ح &آ( 4ا
9اف & ا
!Kهات ا 7 H %'N 4+Xا
!; & 4ا
4
ا
9اف Detrended Series
(6
-3اsن &;%;& w1ك & در 4Lا
! )1و
*61اذا ا;5ج ا&X
6
-4ح ا
!;61ت ا
!;%آI
& 4ا 7N OCا
!; & 4ا
4ا
9اف 7 H %'N
&; 1%C 4ي ا
!آ(ت ا
!4!1
JC -5ر ا
!آ(ت ا
! 4!1آ
;
:7
210
I1 = Median ( d1 , d13 , d 25 ,⋯, d157 )
I 2 = Median ( d 2 , d14 , d 26 ,⋯, d158 )
⋮
I12 = Median ( d12 , d 24 , d 36 ,⋯ , d168 )
.
ات آ:(';
ا1C -6
:7
;
ا آAف
;"ض ه1و
MTB > Read "E:\Mtbwin\milk.dat" c1.
Entering data from file: E:\Mtbwin\milk.dat
168 rows read.
MTB > name c1='MilkProd'
MTB > set c2
DATA> 1:168
DATA> end
MTB > name c2='Time'
MTB > regr c1 1 c2;
SUBC> fits c3.
Regression Analysis
The regression equation is
MilkProd = 612 + 1.69 Time
Predictor Coef StDev T P
Constant 611.682 9.414 64.97 0.000
Time 1.69262 0.09663 17.52 0.000
S = 60.74 R-Sq = 64.9% R-Sq(adj) = 64.7%
Analysis of Variance
Source DF SS MS F
P
Regression 1 1132003 1132003 306.83
0.000
Error 166 612439 3689
Total 167 1744443
211
Unusual Observations
Obs Time MilkProd Fit StDev Fit Residual
St Resid
113 113 942.00 802.95 5.44 139.05
2.30R
125 125 961.00 823.26 6.11 137.74
2.28R
R denotes an observation with a large standardized
residual
MTB > name c3='Trend'
MTB > let c4=c1-c3
MTB > name c4='Detrend'
MTB > Name c5 = 'AVER1'
:*61 و12 4Lك & ا
ر%;& w1;& (6 4
;
ة ا16[
ا7N
MTB > %MA 'Detrend' 12;
SUBC> Center;
SUBC> Averages 'AVER1'.
Executing from file: E:\MTBWIN\MACROS\MA.MAC
Macro is running ... please wait
Moving average
Data Detrend
Length 168.000
NMissing 0
Moving Average
Length: 12
Accuracy Measures
MAPE: 111.68
MAD: 52.36
MSD: 3564.77
ON ا
!ال إ
ا4 ;!
& ا461!
ا4آ%;!
ت ا61;!
ح ا6
212
MTB > let c6=c4-c5
MTB > name c6='DeSeason'
MTB > set c2
DATA> 14(1:12)
DATA> end
MTB > stat c6;
SUBC> by c2;
SUBC> median c7.
MTB > name c7='SeasInx'
Data Display
Row Season SeasInx
1 1 -20.750
2 2 -58.958
3 3 35.625
4 4 50.083
5 5 109.542
6 6 81.292
7 7 33.917
8 8 -10.000
9 9 -52.792
10 10 -50.250
11 11 -79.958
12 12 -44.375
:7+ ا
(
&_ اV& 'ه2L ا7;
ت ا. %
رن اJ
MTB > %Decomp 'MilkProd' 12;
SUBC> Additive ;
SUBC> Start 1.
Executing from file: E:\MTBWIN\MACROS\Decomp.MAC
Macro is running ... please wait
Time Series Decomposition
213
Data MilkProd
Length 168.000
NMissing 0
Trend Line Equation
Yt = 611.682 + 1.69262*t
Seasonal Indices
Period Index
1 -20.1979
2 -58.4062
3 36.1771
4 50.6354
5 110.094
6 81.8437
7 34.4687
8 -9.44792
9 -52.2396
10 -49.6979
11 -79.4063
12 -43.8229
Accuracy of Model
MAPE: 1.583
MAD: 12.088
MSD: 244.406
.;ن2( &; و2JC !O ا
'; ; ا4 رJ!.و
:7
;
آ%Decomp _& (
;[ام ا. ات:('C
1 ف1
MTB > %Decomp 'MilkProd' 12;
SUBC> Additive ;
214
SUBC> Forecasts 12;
SUBC> Start 1.
:ات:(';
ا76"C 7;
وا
Forecasts
Row Period Forecast
1 169 877.54
2 170 841.02
3 171 937.30
4 172 953.45
5 173 1014.60
6 174 988.04
7 175 942.36
8 176 900.13
9 177 859.04
10 178 863.27
11 179 835.25
12 180 872.53
) اةJ
. O
رBة و6"!
; اJ26
;[ام اF. 7&1
اW%
;ج ا9 ات:('C
و: 2!C
_&
(
& ا4 C'
ا
215
ﺍﻟﻔﺼﻞ ﺍﻟﻌﺎﺷﺮ
Using Moving Average Smoothing ك/\ ا0 ; ا0 ا$ 4 و اLا
for Forecasting
'2
آن1
]!N ء6 ا2(C HJ;. f
هات وذK!
اO!;
ك%;!
اw1;!
;[م ا2
m 4Lك & ا
ر%;!
اw1;!
N z1 , z2 , z3 ,… , zn −2 , zn −1 , zn 4'& ز4 ;& & هاتK&
4B "
& اW %2 هاتK!
1
zˆt = ( zt + zt −1 + zt −2 + ⋯ + zt −m+1 ) , t = m, m + 1,..., n
m
أو
1
zˆt = zˆt −1 + ( zt − zt −m ) , t = m, m + 1,..., n
m
. n − m + 1 O!;
" ا. ^(+هات اK!
ان د ا5X
1 ه4]
]
ا4Lك & ا
ر%;!
اw1;!
ن اFN m=3 S
آ1
]!N
1
zˆ3 = ( z3 + z2 + z1 )
3
1 1
zˆ4 = ( z4 + z3 + z2 ) or zˆ4 = zˆ3 + ( z4 − z1 )
3 3
⋮
1 1
zˆn = ( zn + zn−1 + zn −2 ) or zˆn = zˆn −1 + ( z n − zn − 3 )
3 3
ذج1!'
اV(;C هاتK!
;ض ان اb'
ء6 ا2(C HJ;
O!;
اH!"2 e
ى آ73
و
zt = µ + at , at ∼ WN ( 0,σ 2 ) , t = 1, 2,..., n
ن13N
V ( zt ) = σ 2 , ∀t
7
;
.و
σ2
V ( zˆt ) = , t = m, m + 1,..., n
m
216
O!;
ا اA وه4+هات اK!
& اe"P m ـ. Y+ أO'2(C ^(+ة اO!!
هات اK!
أي ان ا
.ء6 اU{C & 6Y&
او1N& آن4 ;!
ا7N w!
أيOI2 ء6
.ةO!!
) اJ
اw1C W' ;'
f
وذ42دN !8 داm A:C :4I5 &
: ك/\ ا0 ام ا:0c$ 4ا
:ك%;!
اw1;!
ا4(J; !
) اJ
|(';! آA:2
zn ( ℓ ) = zˆn −1 , ℓ > 0
:ل-
EMPLOY.MTW H!"
ا4B ا
(
ت & ورH!%
MINITAB 48 5 ا4&%
;[ام اF.
MTB > Retrieve 'E:\Mtbwin\DATA\EMPLOY.MTW'.
اتY;& & ي1%C &ذاI'
MTB > info
Information on the Worksheet
Column Count Name
C1 60 Trade
C2 60 Food
C3 60 Metals
Metals Y;!
هاتK!
ف ;[م ا1
Metals
44.2 44.3 44.4 43.4 42.8 44.3 44.4
44.8 44.4 43.1 42.6 42.4 42.2 41.8
40.1 42.0 42.4 43.1 42.4 43.1 43.2
42.8 43.0 42.8 42.5 42.6 42.3 42.9
43.6 44.7 44.5 45.0 44.8 44.9 45.2
45.2 45.0 45.5 46.2 46.8 47.5 48.3
48.3 49.1 48.9 49.4 50.0 50.0 49.6
49.9 49.6 50.7 50.7 50.9 50.5 51.2
50.7 50.3 49.2 48.1
:هاتK!
@ اA
) ه
217
MTB > TSPlot 'Metals';
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
50
Metals
45
40
Index 10 20 30 40 50 60
ات:('C L1 وm=3 4Lك & ا
ر%;!
اw1;!
;[ام اF. هاتK!
@ اAO
اO!C نs( ا6
:*(J; & )B 6
ـ
MTB > %MA 'Metals' 3;
SUBC> Forecasts 6;
SUBC> Title "Smoothing and Forecasting Metals".
Executing from file: E:\MTBWIN\MACROS\MA.MAC
Moving average
Data Metals
Length 60.0000
NMissing 0
Moving Average
Length: 3
Accuracy Measures
MAPE: 1.55036
218
MAD: 0.70292
MSD: 0.76433
Row Period Forecast Lower Upper
1 61 49.2 47.4865 50.9135
2 62 49.2 47.4865 50.9135
3 63 49.2 47.4865 50.9135
4 64 49.2 47.4865 50.9135
5 65 49.2 47.4865 50.9135
6 66 49.2 47.4865 50.9135
Smoothing and Forecasting Metals
Actual
Predicted
50
Forecast
Actual
Predicted
Forecast
Metals
45
Moving Average
Length: 3
MAPE: 1.55036
MAD: 0.70292
40 MSD: 0.76433
0 10 20 30 40 50 60 _8;'
ا4KB'& :
U
Time
7
%
ا
!]ل ا7N
50.3 + 49.2 + 48.1 147.6
zˆ59 = = = 49.2
3 3
ا ا
!]لA ه7N أوzn +1 , zn + 2 ,..., zn +6 )J
أي4(J; !
ا6 ) ا
ـJ
ات:(';
اA:C
:7
;
آz61 , z62 ,..., z66
z60 (1) = z60 ( 2 ) = ⋯ = z60 ( 6 ) = 49.2
أي zn ( ℓ ) ± 1.96σˆ , ℓ > 0 !ت3
ا W % 95% :('C ;اتN ب%
ر
ـJ! آMSD = 0.76433 4!J
اA{
، 4(J; !
ات ا:(';
) اB V!
[49.2 ± 1.96σˆ ]
219
)J
اV!
95% :('C ;ةN ن13C * وσˆ = 0.8743 ن13N σˆ 2 = 0.76433 أيσ 2
:7 ه4(J; !
ا
49.2 ± 1.96 ( 0.8743) = [ 49.2 ± 1.7135] = [47.4865,50.9135]
:أي
z60+ℓ ∈ [47.4865,50.9135] , ℓ > 0 with probability 0.95
7Cs آMSD W %C :4I5 &
n −1
∑( z i − zˆi )
MSD = σˆ = 2 i =2
n−2
: 2!C
:(';
H`N اO2ر اB و4J.
هات اK!
ا7 7 و5 تL & ا
ر4آ%;& ت61;& (=
.؟4(J; !
) اJ
ا
Running Median ريV\ ا0 ا
4N6;!
ا4!J
N 4N6;!
او اOutliers 4Lهات ا
[رK!
. آ]اU{;2 ك%;!
اw1;!
ا
هاتK!
ا2
S آ1
]!N 4
;;!
ا4آ%;!
ت ا61;!
& اm 7 U:C ة5ا1
ا
z(t)
5 7 3 8 9 6 10 12 1500 11 15
13 18 20
H3K
اO
و
1500
1000
z(t)
500
In d e x 5 10
220
3 4Lك & ا
ر%;& w1;& A{.
M o v in g A v e r a g e
A c tu a l
1500
P re d ic te d
A c tu a l
P re d ic te d
1000
z(t)
M o v in g A v e ra g e
500
L e n gth: 3
M APE: 1081
M AD : 273
0 MSD: 268811
0 5 10 15
T im e
.4N6;!
ا4!J
. )B 4U U 4N تU{C _C'
ك ا%;!
اw1;!
ان ا5X
76 z O!!دي آb
ل ا16
ا
ري ذا اw1
;[م ا2 ت.1"
@ اA هH]& WY;
.4N6;!
) اB. U{;2X يA
وا
W %2 z1 , z2 , z3 ,… , zn −2 , zn −1 , zn هاتK!
j = 2i + 1 ديb
ل ا16
ا
ري ذا اw1
ا
4B "
& ا
zɶt = med ( zt −i ,..., zt ,..., zt +i ) , j = 2i + 1
3 ل16
ري ذا اL w وA{. وzɶt = med ( zt −1 , zt , zt +1 ) 4B "
(^ اC j = 3 4!J
]!N
4J.
هات اK!
15
smoothz(t)
10
In d e x 2 4 6 8 10 12
7
;
اH3K
اO
4JJ%
هات اK!
ن اFN 1500 ~
و15 7 هz9
ـ4JJ%
ا4!J
اS
واذا آ
221
20
15
z(t)
10
Index 5 10
. ; ;'
ا. رنB
222
ﺍﻟﻔﺼﻞ ﺍﻟﺤﺎﺩﻱ ﻋﺸﺮ
Using Single Exponential \ ا0r اL; ا0 ا$ 4 و اLا
: Smoothing for Forecasting
4!2J
) اJ
ن اFN 7
;
. و4!~ اهb
ا
(
تV!L 76"2 ك%;!
اw1;!
ا46ا1. O!;
ا
7X اO!;
ا،%%+ 4!"
ا45'
ن & ا132X B اA وه4]2%
) اJ
آU{;
~ اb
U:C 1
]) اآJ
ا76"2 ~3"
ا
'2
آن1
]!N .O&B V& اoB';C 4! اه6"C يX) اJ
أآ( وا4! أه4Uا5
m 4Lك & ا
ر%;!
اw1;!
N z1 , z2 , z3 ,… , zn −2 , zn −1 , zn 4'& ز4 ;& & هاتK&
4B "
& اW %2 هاتK!
1
zˆt = ( zt + zt −1 + zt −2 + ⋯ + zt −m+1 ) , t = m, m + 1,..., n
m
O;.; آ3!2 7;
وا
1 1 1 1
zˆt = zt + zt −1 + zt −2 + ⋯ + zt −m+1 , t = m, m + 1,..., n
m m m m
1
zˆt = β zt + β zt −1 + β zt −2 + ⋯ + β zt −m+1 , t = m, m + 1,..., n, β =
m
β زن1
~ اb
ا
(
تV!L 76"2 ك%;!
اw1;!
أي ان ا
7
;
آzn ةP%
ا4!J
هات اK!
ُ" ا. V& اoB';C ' ا
(
ت اوزان6 أ1
نsا
st = α zt + α (1 − α ) zt −1 + α (1 − α ) zt −2 + ⋯ , t = 1, 2,..., n, 0 < α < 1
2
O!;
. ! 2& اA وه4J.
) اJ
اV!
اoB';C {وزان. زون1& w1;& 7 هst 4!J
ا
اري3C H3K. W;32 وw (
ا7Xا
st = α zt + (1 − α ) st −1 , t = 1, 2,..., n, s0 = z
ات:(';
اA:Cو
zn ( ℓ ) = sn , ℓ ≥ 1
223
:ل-
EMPLOY.MTW H!"
ا4B ا
(
ت & ورH!%C
MTB > Retrieve 'E:\Mtbwin\DATA\EMPLOY.MTW'.
Metals Y;
ا7N هاتK!
ف
;[م ا1
Metals
44.2 44.3 44.4 43.4 42.8 44.3 44.4
44.8 44.4 43.1 42.6 42.4 42.2 41.8
40.1 42.0 42.4 43.1 42.4 43.1 43.2
42.8 43.0 42.8 42.5 42.6 42.3 42.9
43.6 44.7 44.5 45.0 44.8 44.9 45.2
45.2 45.0 45.5 46.2 46.8 47.5 48.3
48.3 49.1 48.9 49.4 50.0 50.0 49.6
49.9 49.6 50.7 50.7 50.9 50.5 51.2
50.7 50.3 49.2 48.1
:هاتK!
@ اA
) ه
MTB > TSPlot 'Metals';
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
50
Metals
45
40
Index 10 20 30 40 50 60
224
α = 0.2 O!C S.U A{
وw (
ا7X اO!;
;[ام اF. هاتK!
@ اAO
اO!C نs( ا6
:*(J; & )B 6 ات
ـ:('C L1
و
MTB > %SES 'Metals';
SUBC> Weight 0.2;
SUBC> Forecasts 6;
SUBC> Title "Smoothing and Forecasting Metals";
SUBC> Initial 6.
Single Exponential Smoothing
Data Metals
Length 60.0000
NMissing 0
Smoothing Constant
Alpha: 0.2
Accuracy Measures
MAPE: 2.17304
MAD: 1.00189
MSD: 1.45392
Row Period Forecast Lower Upper
1 61 49.7216 47.2670 52.1763
2 62 49.7216 47.2670 52.1763
3 63 49.7216 47.2670 52.1763
4 64 49.7216 47.2670 52.1763
5 65 49.7216 47.2670 52.1763
6 66 49.7216 47.2670 52.1763
225
Smoothing and Forecasting Metals
Actual
Predicted
50 Forecast
Actual
Predicted
Metals
Forecast
45 Smoothing Constant
Alpha: 0.200
MAPE: 2.17304
MAD: 1.00189
40 MSD: 1.45392
0 10 20 30 40 50 60
Time
_8;'
ا4KB'& :
U
α = 0.2 O!C S.]. z1 , z2 ,… , zn −1 , zn −2 هاتK!
w (
ا7X اO!;
اW %2 -1
:42ار3;
ا4B "
& ا
si = α zi + (1 − α ) si −1 , i = 1,2,..., n
W %C 7;
واs0 4
وX ا4!J
ا7
;ج ا%
ة اO!!
) اJ
ب ا%
42ار3;
ا4B "
(أ ا73
w1;!
42 & وs0 VP و7 هO&[; ' 7;
ق وا6
@ اA ه5 أ،"ة =ق.
m
∑z i
'
]& 7bN s0 = i =1
, m = 6 ( or n, if n<6 )
m
44.2 + 44.3 + 44.4 + 43.4 + 42.8 + 44.3
s0 = = 43.9
6
ن132 7
;
.و
s1 = α z1 + (1 − α ) s0 = 0.2 ( 44.2 ) + 0.8 ( 43.9 ) = 8.84 + 35.12 = 43.96
s2 = α z2 + (1 − α ) s1 = 0.2 ( 44.3) + 0.8 ( 43.96 ) = 8.86 + 35.168 = 44.028
:7
;
';_ اN هةK& , 7;5 !; اA3وه
Time Metals SMOO1 FITS1 RESI1
1 44.2 43.9600 43.9000 0.30000
2 44.3 44.0280 43.9600 0.34000
3 44.4 44.1024 44.0280 0.37200
4 43.4 43.9619 44.1024 -0.70240
226
5 42.8 43.7295 43.9619 -1.16192
6 44.3 43.8436 43.7295 0.57046
7 44.4 43.9549 43.8436 0.55637
8 44.8 44.1239 43.9549 0.84510
9 44.4 44.1791 44.1239 0.27608
10 43.1 43.9633 44.1791 -1.07914
11 42.6 43.6906 43.9633 -1.36331
12 42.4 43.4325 43.6906 -1.29065
13 42.2 43.1860 43.4325 -1.23252
14 41.8 42.9088 43.1860 -1.38601
15 40.1 42.3470 42.9088 -2.80881
16 42.0 42.2776 42.3470 -0.34705
17 42.4 42.3021 42.2776 0.12236
18 43.1 42.4617 42.3021 0.79789
19 42.4 42.4494 42.4617 -0.06169
20 43.1 42.5795 42.4494 0.65065
21 43.2 42.7036 42.5795 0.62052
22 42.8 42.7229 42.7036 0.09642
23 43.0 42.7783 42.7229 0.27713
24 42.8 42.7826 42.7783 0.02171
25 42.5 42.7261 42.7826 -0.28264
26 42.6 42.7009 42.7261 -0.12611
27 42.3 42.6207 42.7009 -0.40089
28 42.9 42.6766 42.6207 0.27929
29 43.6 42.8613 42.6766 0.92343
30 44.7 43.2290 42.8613 1.83875
31 44.5 43.4832 43.2290 1.27100
32 45.0 43.7866 43.4832 1.51680
33 44.8 43.9892 43.7866 1.01344
34 44.9 44.1714 43.9892 0.91075
35 45.2 44.3771 44.1714 1.02860
36 45.2 44.5417 44.3771 0.82288
37 45.0 44.6334 44.5417 0.45830
38 45.5 44.8067 44.6334 0.86664
227
39 46.2 45.0853 44.8067 1.39331
40 46.8 45.4283 45.0853 1.71465
41 47.5 45.8426 45.4283 2.07172
42 48.3 46.3341 45.8426 2.45738
43 48.3 46.7273 46.3341 1.96590
44 49.1 47.2018 46.7273 2.37272
45 48.9 47.5415 47.2018 1.69818
46 49.4 47.9132 47.5415 1.85854
47 50.0 48.3305 47.9132 2.08683
48 50.0 48.6644 48.3305 1.66947
49 49.6 48.8515 48.6644 0.93557
50 49.9 49.0612 48.8515 1.04846
51 49.6 49.1690 49.0612 0.53877
52 50.7 49.4752 49.1690 1.53101
53 50.7 49.7202 49.4752 1.22481
54 50.9 49.9561 49.7202 1.17985
55 50.5 50.0649 49.9561 0.54388
56 51.2 50.2919 50.0649 1.13510
57 50.7 50.3735 50.2919 0.40808
58 50.3 50.3588 50.3735 -0.07353
59 49.2 50.1271 50.3588 -1.15883
60 48.1 49.7216 50.1271 -2.02706
FITS1 ~&[
د ا1!"
اsi , i = 1,2,...,60 ة أيO!!
) اJ
ى ا1%2 SMOO1 V.د ا
ا1!"
ا
ء6ي ا1%2 RESI1 ~&[
د ا1!"
اzˆi = si −1 , i = 1, 2,...,60 أي4J(6!
) اJ
ى ا1%2
ei = zi − zˆi , i = 1, 2,...,60 ( أيResiduals 7Bا1(
)ا
:ة أيO!& 4!B , 4(J; !
) اJ
|(';! آA:2 -2
zn ( ℓ ) = sn , ℓ > 0
7
%
ا
!]ل ا7bN
z60 ( ℓ ) = 49.7216, ℓ > 0
228
ا ا
!]لA ه7N أوzn +1 , zn +2 ,..., zn +6 )J
أي4(J; !
ا6 ) ا
ـJ
ات:(';
اA:C
:7
;
آz61 , z62 ,..., z66
z60 (1) = z60 ( 2 ) = ⋯ = z60 ( 6 ) = 49.7216
أي zn ( ℓ ) ± 1.96σˆ , ℓ > 0 !ت3
ا W % 95% :('C ;اتN ب%
-3
MSD = 1.45392 4!J
اA{ ، 4(J; !
ات ا:(';
) اB V!
[49.7216 ± 1.96σˆ ]
95% :('C ;ةN ن13C * وσˆ = 1.205786 ن13N σˆ 2 = 1.45392 أيσ 2 ر
ـJ!آ
:7 ه4(J; !
) اJ
اV!
49.7216 ± 1.96 (1.205786 ) = [49.7216 ± 2.3633] = [47.35826,52.08494]
:أي
z60+ℓ ∈ [47.3582,50.0849] , ℓ > 0 with probability 0.95
7Cs آMSD W %C :4I5 &
n
∑( z i − zˆi )
MSD = σˆ = 2 i =1
n −1
: 2!C
H`N اO2ر اB وα = 0.3,0.4,0.5 &[; & 4J.
هات اK!
ا7 w . 7 اO!C (=
.؟4(J; !
) اJ
ا:(';
229
ﺍﻟﻔﺼﻞ ﺍﻟﺜﺎﻧﻲ ﻋﺸﺮ
Using Double Exponential ادوج0r اL; ا0 ا$ 4 و اLا
: Smoothing for Forecasting
:Brown’s Method اون. 4J2= :Xأو
:7
;
اL1 0 < α < 1 O!C S.]
وz1 , z2 ,… , zn −1 , zn −2 هاتK!
st( ) = α zt + (1 − α ) st(−1) , t = 1,2,..., n
1 1
اA ه4L در7
& اC (1) ( وw (
ا7X اO!;
اI )اst( ) = st w . 7 اO!C st( ) Q5
1 1
O!;
ا
st( ) = α st( ) + (1 − α ) st(−1) , t = 1, 2,..., n
2 1 2
O!;
ا اA ه4L در7
& اC ( 2 ) &دوج و7 اO!C st( ) Q5
2
at = 2 st( ) − st( ) , t = 1, 2,..., n
1 2
α
bt =
1−α
1
( 2
)
st( ) − st( ) , t = 1, 2,..., n
4
& ا
!"د4J(6!
) اJ
اW %C
zˆt = at + bt t , t = 1, 2,..., n
& zn +ℓ , ℓ > 0 4(J; !
) اJ
ات:(';
اW %Cو
zn ( ℓ ) = an + bn ℓ, ℓ > 0
4J.
ت اB "
& ا: s0( ) وs0( ) 4
وX) اJ
ب ا5
2 1
1−α
s0( ) = a0 −
1
b0
α
1−α
s0( ) = a0 − 2
2
b0
α
ن132 وzt = α + β t + et , t = 1,2,..., n &
ا7 هاتK!
ار ا%
F. b0 وa0 L1
b0 = βˆ وa0 = αˆ
230
:Holt’s Method S
1 ه4J2= :
U
:7
;
اL1
0 < γ < 1 و0 < α < 1 O!C 7;.]
وz1 , z2 ,… , zn −1 , zn −2 هاتK!
st = α zt + (1 − α )( st −1 + bt −1 ) , t = 1, 2,..., n
bt = γ ( st − st −1 ) + (1 − γ ) bt −1 , t = 1, 2,..., n
& 4J(6!
) اJ
اW %
zˆt = st + bt t , t = 1, 2,..., n
& 4(J; !
) اJ
ات:(';
وا
zn ( ℓ ) = sn + bn ℓ, ℓ > 0
& b0 وs0 4
وX) اJ
اW %
s0 = z1
b0 = z2 − z1 or
( z2 − z1 ) + ( z3 − z2 ) ( z3 − z1 )
b0 = = or
2 2
( z − z ) + ( z3 − z2 ) + ( z4 − z3 ) ( z4 − z1 )
b0 = 2 1 =
3 3
:ل-
EMPLOY.MTW H!"
ا4B ا
(
ت & ورH!%C
MTB > Retrieve 'E:\Mtbwin\DATA\EMPLOY.MTW'.
Metals Y;
ا7N هاتK!
ف
;[م ا1
Metals
44.2 44.3 44.4 43.4 42.8 44.3 44.4
44.8 44.4 43.1 42.6 42.4 42.2 41.8
40.1 42.0 42.4 43.1 42.4 43.1 43.2
42.8 43.0 42.8 42.5 42.6 42.3 42.9
43.6 44.7 44.5 45.0 44.8 44.9 45.2
45.2 45.0 45.5 46.2 46.8 47.5 48.3
48.3 49.1 48.9 49.4 50.0 50.0 49.6
49.9 49.6 50.7 50.7 50.9 50.5 51.2
50.7 50.3 49.2 48.1
231
:هاتK!
@ اA
) ه
MTB > TSPlot 'Metals';
SUBC> Index;
SUBC> TDisplay 11;
SUBC> Symbol;
SUBC> Connect.
50
Metals
45
40
Index 10 20 30 40 50 60
نs ;[م ا،اون. 4J26. ا
!دوج7X اO!;
;[ام اF. هاتK!
@ اAO
اO!C نs( ا6
هA{ ف1 42{وزان &; و. WEIGHT 7b
& اX اV& %DES (Macro) _&(
ا
0.2
MTB > %DES 'Metals';
SUBC> Weight 0.2 0.2;
SUBC> Forecasts 6;
SUBC> Title "Brown's Double Exponential Smoothing";
SUBC> Table.
Double Exponential Smoothing
Data Metals
Length 60.0000
232
NMissing 0
Smoothing Constants
Alpha (level): 0.2
Gamma (trend): 0.2
Accuracy Measures
MAPE: 2.16187
MAD: 0.97032
MSD: 1.62936
Time Metals Smooth Predict Error
1 44.2 41.7739 41.1674 3.03257
2 44.3 42.4976 42.0470 2.25303
3 44.4 43.1686 42.8607 1.53927
4 43.4 43.5546 43.5933 -0.19330
5 42.8 43.7373 43.9716 -1.17163
6 44.3 44.1459 44.1074 0.19257
7 44.4 44.4990 44.5238 -0.12377
8 44.8 44.8575 44.8719 -0.07189
9 44.4 45.0620 45.2275 -0.82751
10 43.1 44.9391 45.3989 -2.29891
11 42.6 44.6673 45.1841 -2.58407
12 42.4 44.3271 44.8088 -2.40884
13 42.2 43.9378 44.3723 -2.17229
14 41.8 43.4769 43.8962 -2.09617
15 40.1 42.7011 43.3514 -3.25142
16 42.0 42.3565 42.4456 -0.44557
17 42.4 42.1465 42.0831 0.31694
18 43.1 42.1286 41.8857 1.21426
19 42.4 42.0132 41.9164 0.48355
233
20 43.1 42.0763 41.8204 1.27964
21 43.2 42.1877 41.9347 1.26533
22 42.8 42.2374 42.0967 0.70327
23 43.0 42.3396 42.1745 0.82549
24 42.8 42.4078 42.3098 0.49024
25 42.5 42.4181 42.3976 0.10244
26 42.6 42.4495 42.4119 0.18809
27 42.3 42.4207 42.4509 -0.15090
28 42.9 42.5129 42.4161 0.48394
29 43.6 42.7420 42.5276 1.07245
30 44.7 43.1797 42.7996 1.90036
31 44.5 43.5507 43.3133 1.18668
32 45.0 43.9854 43.7317 1.26826
33 44.8 44.3338 44.2172 0.58280
34 44.9 44.6511 44.5889 0.31112
35 45.2 44.9749 44.9187 0.28133
36 45.2 45.2430 45.2538 -0.05376
37 45.0 45.4157 45.5197 -0.51967
38 45.5 45.6373 45.6716 -0.17162
39 46.2 45.9491 45.8863 0.31368
40 46.8 46.3285 46.2106 0.58938
41 47.5 46.7909 46.6136 0.88637
42 48.3 47.3492 47.1115 1.18850
43 48.3 47.8339 47.7173 0.58266
44 49.1 48.4002 48.2253 0.87469
45 48.9 48.8413 48.8267 0.07332
46 49.4 49.2966 49.2707 0.12930
47 50.0 49.7849 49.7311 0.26890
48 50.0 50.1841 50.2302 -0.23017
49 49.6 50.4162 50.6202 -1.02022
50 49.9 50.6292 50.8114 -0.91145
51 49.6 50.7104 50.9880 -1.38797
52 50.7 50.9509 51.0137 -0.31368
53 50.7 51.1334 51.2417 -0.54169
234
54 50.9 51.3019 51.4024 -0.50244
55 50.5 51.3407 51.5509 -1.05093
56 51.2 51.4782 51.5477 -0.34770
57 50.7 51.4770 51.6712 -0.97120
58 50.3 51.3649 51.6311 -1.33115
59 49.2 51.0127 51.4659 -2.26587
60 48.1 50.4384 51.0230 -2.92300
Row Period Forecast Lower Upper
1 61 50.3318 47.9545 52.7091
2 62 50.2252 47.7984 52.6520
3 63 50.1186 47.6384 52.5987
4 64 50.0120 47.4749 52.5490
5 65 49.9054 47.3080 52.5027
6 66 49.7988 47.1381 52.4594
Brown's Double Exponential Smoothing
Actual
Predicted
Forecast
50 Actual
Predicted
Forecast
Metals
45 Smoothing Constants
Alpha (level): 0.200
Gamma (trend):0.200
MAPE: 2.16187
MAD: 0.97032
40 MSD: 1.62936
0 10 20 30 40 50 60
Time
235
: b0 وa0 د2إ
MTB > set c4
DATA> 1:60
DATA> end
MTB > regr c3 1 c4
Regression Analysis
The regression equation is
Metals = 41.0 + 0.152 C4
W % O'& وb0 = 0.152 وa0 = 41.0 إذا
1−α 0.8
s0( ) = a0 −
1
b0 = 41.0 − ( 0.152 ) = 41.608
α 0.2
1−α 0.8
s0( ( 0.152 ) = 42.216
2)
= a0 − 2 b0 = 41.0 − 2
α 0.2
s1( ) = ( 0.2 )( 44.2 ) + ( 0.8 )( 41.608 ) = 42.1264
1
s1( ) = ( 0.2 )( 42.1264 ) + ( 0.8 )( 42.216 ) = 42.19808
2
a1 = ( 2 )( 42.1264 ) − 42.19808 = 42.05472
( 0.2 )
b1 = ( 42.19808 − 42.05472 ) = 0.03584
( 0.8 )
zˆ1 = 42.05472 + ( 0.03584 )(1) = 42.09056
… r
ا اA3وه
.4J.
ا4]& ا7N ! آMSD ;[امF. :(';
;ات اN W %C
:ل-
WEIGHT 7b
& اX اV& %DES (Macro) _&(
ن اs
;[م اS
1 ه4J2= (6;
γ = 0.3 وα = 0.2 A{ ف1 4b;[& {وزان.
MTB > RETR 'E:\Mtbwin\DATA\EMPLOY.MTW'.
Retrieving worksheet from file: E:\Mtbwin\DATA\EMPLOY.MTW
Worksheet was saved on 6/ 5/1996
236
MTB > %DES 'Metals';
SUBC> Weight 0.2 0.3;
SUBC> Forecasts 6;
SUBC> Title "Holt's Double Exponential Smoothing";
SUBC> Table.
Double Exponential Smoothing
Data Metals
Length 60.0000
NMissing 0
Smoothing Constants
Alpha (level): 0.2
Gamma (trend): 0.3
Accuracy Measures
MAPE: 2.15656
MAD: 0.96328
MSD: 1.56274
Time Metals Smooth Predict Error
1 44.2 41.7739 41.1674 3.03257
2 44.3 42.5461 42.1076 2.19238
3 44.4 43.2891 43.0113 1.38868
4 43.4 43.7501 43.8376 -0.43760
5 42.8 43.9779 44.2724 -1.47237
6 44.3 44.3895 44.4118 -0.11184
7 44.4 44.7334 44.8167 -0.41671
8 44.8 45.0685 45.1356 -0.33560
9 44.4 45.2405 45.4506 -1.05057
237
10 43.1 45.0676 45.5595 -2.45952
11 42.6 44.7113 45.2391 -2.63911
12 42.4 44.2595 44.7244 -2.32443
13 42.2 43.7466 44.1332 -1.93322
14 41.8 43.1634 43.5043 -1.70426
15 40.1 42.2751 42.8188 -2.71884
16 42.0 41.8139 41.7674 0.23263
17 42.4 41.5361 41.3202 1.07985
18 43.1 41.5057 41.1072 1.99283
19 42.4 41.4371 41.1964 1.20365
20 43.1 41.5799 41.1999 1.90008
21 43.2 41.8054 41.4568 1.74322
22 42.8 41.9895 41.7869 1.01314
23 43.0 42.2254 42.0317 0.96829
24 42.8 42.4205 42.3257 0.47431
25 42.5 42.5395 42.5493 -0.04933
26 42.6 42.6522 42.6653 -0.06528
27 42.3 42.6793 42.7741 -0.47413
28 42.9 42.7982 42.7728 0.12724
29 43.6 43.0394 42.8993 0.70070
30 44.7 43.4861 43.1826 1.51743
31 44.5 43.8762 43.7202 0.77977
32 45.0 44.3257 44.1571 0.84285
33 44.8 44.6858 44.6573 0.14275
34 44.9 45.0007 45.0259 -0.12590
35 45.2 45.3066 45.3333 -0.13327
36 45.2 45.5449 45.6312 -0.43116
37 45.0 45.6749 45.8436 -0.84361
38 45.5 45.8384 45.9230 -0.42295
39 46.2 46.0888 46.0610 0.13895
40 46.8 46.4159 46.3199 0.48014
41 47.5 46.8406 46.6757 0.82428
42 48.3 47.3799 47.1499 1.15014
43 48.3 47.8665 47.7582 0.54181
238
44 49.1 48.4419 48.2774 0.82265
45 48.9 48.9016 48.9020 -0.00205
46 49.4 49.3693 49.3617 0.03832
47 50.0 49.8653 49.8317 0.16832
48 50.0 50.2702 50.3378 -0.33779
49 49.6 50.4979 50.7224 -1.12240
50 49.9 50.6862 50.8827 -0.98275
51 49.6 50.7296 51.0121 -1.41206
52 50.7 50.9166 50.9708 -0.27079
53 50.7 51.0532 51.1415 -0.44152
54 50.9 51.1813 51.2516 -0.35162
55 50.5 51.1869 51.3586 -0.85860
56 51.2 51.2901 51.3127 -0.11267
57 50.7 51.2673 51.4092 -0.70916
58 50.3 51.1350 51.3438 -1.04381
59 49.2 50.7591 51.1489 -1.94889
60 48.1 50.1448 50.6560 -2.55603
Row Period Forecast Lower Upper
1 61 49.8884 47.5283 52.2484
2 62 49.6319 47.1597 52.1041
3 63 49.3755 46.7803 51.9707
4 64 49.1190 46.3915 51.8466
5 65 48.8626 45.9946 51.7306
6 66 48.6061 45.5908 51.6215
239
Holt's Double Exponential Smoothing
Actual
Predicted
Forecast
50
Actual
Predicted
Forecast
Metals
45 Smoothing Constants
Alpha (level): 0.200
Gamma (trend):0.300
MAPE: 2.15656
MAD: 0.96328
40 MSD: 1.56274
0 10 20 30 40 50 60
Time
.6C مVB1;2 :4I5 & ) 2و2 )J
"\ ا. V(;;. 4J.
ت ا. %
ا4%+ & J%C : 2!C
7N f
A وآ4(%
ا4
s واW%
ا. اادH]!C 2= ; ف9 f
!& وذC ت. %
ا
(!O'& H ذاآت آ7N مB ار2[C
240
ﺍﻟﻔﺼﻞ ﺍﻟﺜﺎﻟﺚ ﻋﺸﺮ
a
V ا0 ت ا,
ز7
وi ;0 ا$ 4] و ا,- ا0r اLا
Triple Exponential Smoothing: Winters' Three-Parameter Trend
and Seasonality Smoothing Method
&ى4!1!
اه ا1I
اH%;
Vb'C X H b
ا اA ه7N در'ه7;
ا4J.
ق ا6
اV!L
Winters' trend and و
;ز4J2= وDecomposition Method f3b;
ا4J2=
' هOP";
ف1 7;
اseasonal smoothing
4N '!
ا4!1!
و
;ز
!; ت ا4J2=
4B "
. ا آO!C هاتK!
اO!C :Xأو
zt
st = α + (1 − α )( st −1 + bt −1 ) , t = 1,2,..., n
St − s
اف9 اO!C :
U
bt = γ ( st − st −1 ) + (1 − γ ) bt −1 , t = 1,2,..., n
4!1!
اO!C :]
U
zt
St = β + (1 − β ) St −s , t = 1, 2,..., n
st
4!1!
دورة ا7 هs وi &
' ا4!1!
ا4( ا
!آ7 هSi Q5
4B "
. 76"C 4J(6!
) اJ
ا
zˆt = ( st + bt t ) St −s , t = 1, 2,..., n
4B "
ات & ا:(';
وا
zn ( ℓ ) = ( sn + bn ℓ ) Sn −s +ℓ , ℓ > 0
ارز&ت1[. W %C 4
وX) اJ
ان اQ5 42ت ا
و. %
. و
;ز4J2= V(;C اL W"
& ا
.W%
& ا4L[!
_ ا8;'
. 7b;3
ه' وOP";
اAO
وW%
;[ام اF. 46 z
:ل-
EMPLOY.MTW H!"
ا4B ا
(
ت & ورH!%C
241
MTB > Retrieve 'E:\Mtbwin\DATA\EMPLOY.MTW'.
Food Y;!
ا7N هاتK!
ف
;[م ا1
Food
53.5 53.0 53.2 52.5 53.4 56.5 65.3 70.7 66.9 58.2 55.3 53.4
52.1 51.5 51.5 52.4 53.3 55.5 64.2 69.6 69.3 58.5 55.3 53.6
52.3 51.5 51.7 51.5 52.2 57.1 63.6 68.8 68.9 60.1 55.6 53.9
53.3 53.1 53.5 53.5 53.9 57.1 64.7 69.4 70.3 62.6 57.9 55.8
54.8 54.2 54.6 54.3 54.8 58.1 68.1 73.3 75.5 66.4 60.5 57.7
هاتK!
و
) ا
75
70
65
Food
60
55
50
Index 10 20 30 40 50 60
:7
;
و
;ز آ4J2= نs( ا6
12 ورة. 4!1& هةI
ان ا5
Additive Model 7NP9ذج ا1!'
:Xأو
zt = bt + St + et , t = 1, 2,..., n
MTB > %Wintadd 'Food' 12;
SUBC> Weight 0.2 0.2 0.2;
SUBC> Forecasts 12;
SUBC> Title "Wintrs' Trend and Seasonal Smoothing";
242
SUBC> Table.
Winters' additive model
Data Food
Length 60.0000
NMissing 0
Smoothing Constants
Alpha (level): 0.2
Gamma (trend): 0.2
Delta (seasonal): 0.2
Accuracy Measures
MAPE: 1.94769
MAD: 1.15100
MSD: 2.66711
Time Food Smooth Predict Error
1 53.5 48.7755 49.4303 4.06965
2 53.0 49.6020 50.4197 2.58027
3 53.2 51.0736 51.9944 1.20556
4 52.5 52.0733 53.0424 -0.54244
5 53.4 53.5117 54.4591 -1.05914
6 56.5 57.4851 58.3901 -1.89013
7 65.3 66.2299 67.0593 -1.75932
8 70.7 71.7852 72.5443 -1.84430
9 66.9 71.8932 72.5785 -5.67851
10 58.2 62.3206 62.7787 -4.57874
11 55.3 57.5208 57.7958 -2.49577
12 53.4 55.1544 55.3296 -1.92957
13 52.1 55.0393 55.1373 -3.03734
14 51.5 53.6493 53.6258 -2.12584
243
15 51.5 53.1185 53.0100 -1.50996
16 52.4 52.2661 52.0971 0.30287
17 53.3 52.6528 52.4960 0.80401
18 55.5 55.7616 55.6369 -0.13695
19 64.2 63.8483 63.7181 0.48187
20 69.6 68.8787 68.7678 0.83218
21 69.3 68.0386 67.9610 1.33902
22 58.5 59.2825 59.2585 -0.75851
23 55.3 55.0979 55.0435 0.25650
24 53.6 53.0432 52.9991 0.60092
25 52.3 53.0377 53.0177 -0.71765
26 51.5 52.1394 52.0907 -0.59067
27 51.7 51.9889 51.9165 -0.21651
28 51.5 51.7214 51.6403 -0.14031
29 52.2 52.1875 52.1009 0.09913
30 57.1 55.0750 54.9923 2.10774
31 63.6 63.7515 63.7531 -0.15314
32 68.8 68.8427 68.8382 -0.03822
33 68.9 68.0160 68.0099 0.89007
34 60.1 58.9061 58.9356 1.16436
35 55.6 55.3220 55.3981 0.20190
36 53.9 53.4419 53.5262 0.37384
37 53.3 53.3084 53.4076 -0.10757
38 53.1 52.6717 52.7666 0.33345
39 53.5 52.9095 53.0177 0.48233
40 53.5 52.9745 53.1020 0.39803
41 53.9 53.7952 53.9386 -0.03858
42 57.1 57.2065 57.3484 -0.24838
43 64.7 65.2747 65.4066 -0.70661
44 69.4 70.4039 70.5076 -1.10758
45 70.3 69.6200 69.6794 0.62065
46 62.6 60.5655 60.6497 1.95031
47 57.9 57.0392 57.2014 0.69858
48 55.8 55.3721 55.5623 0.23773
244
49 54.8 55.2403 55.4399 -0.63993
50 54.2 54.6681 54.8422 -0.64220
51 54.6 54.8138 54.9622 -0.36218
52 54.3 54.7366 54.8705 -0.57048
53 54.8 55.3001 55.4112 -0.61119
54 58.1 58.5310 58.6176 -0.51765
55 68.1 66.4168 66.4827 1.61731
56 73.3 71.8806 72.0112 1.28878
57 75.5 71.8794 72.0616 3.43843
58 66.4 63.7240 64.0437 2.35629
59 60.5 60.3141 60.7281 -0.22810
60 57.7 58.6397 59.0446 -1.34455
Row Period Forecast Lower Upper
1 61 58.6167 55.7968 61.4366
2 62 58.3236 55.4449 61.2023
3 63 58.8195 55.8775 61.7614
4 64 58.9840 55.9746 61.9935
5 65 59.8723 56.7913 62.9532
6 66 63.4804 60.3243 66.6365
7 67 72.0757 68.8410 75.3104
8 68 77.4486 74.1321 80.7651
9 69 77.7540 74.3528 81.1552
10 70 68.9067 65.4180 72.3954
11 71 64.6434 61.0647 68.2221
12 72 62.7731 59.1020 66.4441
245
Wintrs' Trend and Seasonal Smoothing
Actual
80 Predicted
Forecast
Actual
Predicted
70
Forecast
Food
Smoothing Constants
Alpha (level): 0.200
60
Gamma (trend):0.200
Delta (season):0.200
MAPE: 1.94769
50 MAD: 1.15100
MSD: 2.66711
0 10 20 30 40 50 60 70
Time
Multiplicative Model 7b`;
ذج ا1!'
: U
zt = bt St + et , t = 1, 2,..., n
MTB > %Wintmult 'Food' 12;
SUBC> Weight 0.2 0.2 0.2;
SUBC> Forecasts 12;
SUBC> Title "Winters' Trend and Seasonal Smoothing";
SUBC> Table.
Winters' multiplicative model
Data Food
Length 60.0000
NMissing 0
Smoothing Constants
Alpha (level): 0.2
Gamma (trend): 0.2
Delta (seasonal): 0.2
Accuracy Measures
MAPE: 1.88377
MAD: 1.12068
MSD: 2.86696
246
Time Food Smooth Predict Error
1 53.5 48.7870 49.3853 4.11470
2 53.0 49.6755 50.4303 2.56966
3 53.2 51.1521 52.0132 1.18677
4 52.5 52.1675 53.0746 -0.57458
5 53.4 53.6181 54.5132 -1.11323
6 56.5 57.6509 58.5541 -2.05414
7 65.3 66.6199 67.5607 -2.26072
8 70.7 72.4105 73.3280 -2.62800
9 66.9 72.5679 73.3777 -6.47768
10 58.2 62.7837 63.2634 -5.06337
11 55.3 57.9154 58.1732 -2.87320
12 53.4 55.5108 55.6485 -2.24849
13 52.1 54.4920 54.5392 -2.43920
14 51.5 53.2117 53.1621 -1.66212
15 51.5 52.8118 52.6957 -1.19573
16 52.4 52.0929 51.9302 0.46985
17 53.3 52.5894 52.4439 0.85611
18 55.5 55.7388 55.6209 -0.12087
19 64.2 63.7189 63.5782 0.62178
20 69.6 68.7087 68.5838 1.01617
21 69.3 67.9722 67.8890 1.41104
22 58.5 59.4594 59.4361 -0.93606
23 55.3 55.4037 55.3468 -0.04680
24 53.6 53.4103 53.3536 0.24639
25 52.3 52.6818 52.6356 -0.33562
26 51.5 51.8659 51.8071 -0.30705
27 51.7 51.8002 51.7290 -0.02902
28 51.5 51.6271 51.5549 -0.05492
29 52.2 52.1643 52.0890 0.11103
247
30 57.1 55.0424 54.9676 2.13244
31 63.6 63.6079 63.6199 -0.01988
32 68.8 68.6702 68.6823 0.11774
33 68.9 67.9561 67.9727 0.92727
34 60.1 59.1021 59.1487 0.95133
35 55.6 55.6210 55.7003 -0.10032
36 53.9 53.7881 53.8609 0.03912
37 53.3 53.0479 53.1211 0.17892
38 53.1 52.4502 52.5294 0.57055
39 53.5 52.7444 52.8467 0.65329
40 53.5 52.8747 53.0029 0.49714
41 53.9 53.7689 53.9188 -0.01879
42 57.1 57.2790 57.4374 -0.33743
43 64.7 65.4702 65.6357 -0.93567
44 69.4 70.6713 70.8095 -1.40954
45 70.3 69.8908 69.9719 0.32815
46 62.6 60.7552 60.8370 1.76302
47 57.9 57.1925 57.3348 0.56523
48 55.8 55.5181 55.6775 0.12253
49 54.8 54.8764 55.0383 -0.23826
50 54.2 54.3244 54.4749 -0.27486
51 54.6 54.5372 54.6769 -0.07694
52 54.3 54.5298 54.6661 -0.36612
53 54.8 55.1925 55.3155 -0.51551
54 58.1 58.6054 58.7141 -0.61410
55 68.1 66.7739 66.8698 1.23016
56 73.3 72.4056 72.5622 0.73784
57 75.5 72.3385 72.5236 2.97638
58 66.4 63.6729 63.9378 2.46217
59 60.5 60.0395 60.3781 0.12191
60 57.7 58.3023 58.6338 -0.93381
Row Period Forecast Lower Upper
248
1 61 57.8102 55.0645 60.5558
2 62 57.3892 54.5864 60.1921
3 63 57.8332 54.9687 60.6977
4 64 57.9307 55.0005 60.8609
5 65 58.8311 55.8313 61.8309
6 66 62.7415 59.6686 65.8145
7 67 72.1849 69.0354 75.3344
8 68 78.1507 74.9215 81.3798
9 69 78.5092 75.1976 81.8208
10 70 68.6689 65.2721 72.0657
11 71 63.9258 60.4414 67.4103
12 72 61.8189 58.2446 65.3933
Winters' Trend and Seasonal Smoothing
Actual
80 Predicted
Forecast
Actual
Predicted
70 Forecast
Food
Smoothing Constants
Alpha (level): 0.200
60 Gamma (trend):0.200
Delta (season):0.200
MAPE: 1.88377
50 MAD: 1.12068
MSD: 2.86696
0 10 20 30 40 50 60 70
Time
:تI5 &
S.U γ و73
اO!;
اS.U α 7 &"
) ه4U U )B إ;ر7
;ج ا%
و
;ز4J2= (6;
4U U 7N Optimization 4`N أ4! @A وه4!1!
اO!C S.U β اف و9 اO!C
;[امF. 8JC O. 5 ا
! ;[م78 59`ء ا
!"
) ( إ& أن ;ك
( &_ اN ) "د.ا
.)J
اf;. _& (
&اد اF. % م1J ا
( &_ أوH دا4'(& 46 z ارز&ت1
Q
]
اz &ة &"! وH آ7N S(U . α = γ = β = 0.2 A ا4J.
ا4]& ا7N : 2!C
MSD HB أ7 H %C 7;5
249
:4 ذج7 ل ء-
MINITAB _& (
ا
( ت41! & & CPI.MTW H!"
ا4Bف ;[م ور1
MTB > Retrieve 'C:\MTBWIN\STUDENT9\CPI.MTW'.
CPIChange Y;!
اA{
ف1
CPIChnge
1.7 1.0 1.0 1.3 1.3 1.6 2.9
3.1 4.2 5.5 5.7 4.4 3.2 6.2
11.0 9.1 5.8 6.5 7.6 11.3 13.5
10.3 6.2 3.2 4.3 3.6 1.9 3.6
4.1 4.8 5.4 4.2 3.0
4'"
ا48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6.;
اL1
و4 ;!
) ا
MTB > TSPlot
14
12
10
CPIChnge
0
Index 5 10 15 20 25 30
MTB > %acf c2
Autocorrelation Function for CPIChnge
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8
Lag Corr T LBQ Lag Corr T LBQ
1 0.79 4.53 22.42 8 -0.16 -0.51 41.23
2 0.46 1.77 30.32
3 0.29 1.02 33.59
4 0.26 0.88 36.24
5 0.26 0.86 38.97
6 0.16 0.52 40.05
7 -0.02 -0.06 40.07
250
MTB > %pacf c2
Partial Autocorrelation Function for CPIChnge
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8
Lag PAC T Lag PAC T
1 0.79 4.53 8 -0.09 -0.52
2 -0.42 -2.44
3 0.35 2.01
4 -0.04 -0.24
5 0.08 0.43
6 -0.28 -1.62
7 -0.05 -0.29
ARMA(1,1) ذج1!'
ن ا132 B 4'"
ا48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6.;
& ا
!ط ا
4(J; !
) اJ
ات:('C 5
1C;ح وJ!
ذج ا1!'
( ا6C 4
;
اوا& ا،4 ;!
( ا6'2
MTB > arima 1 0 1 c2;
SUBC> fore 5 c3 c4 c5;
SUBC> gser;
SUBC> gacf;
SUBC> gpacf;
SUBC> ghist;
SUBC> gnormal.
ARIMA Model
ARIMA model for CPIChnge
Estimates at each iteration
Iteration SSE Parameters
0 323.251 0.100 0.100 4.522
1 200.616 0.250 -0.050 3.745
2 182.146 0.184 -0.200 4.067
3 163.067 0.135 -0.350 4.308
251
4 142.864 0.107 -0.500 4.434
5 121.402 0.111 -0.650 4.407
6 99.668 0.150 -0.800 4.197
7 77.036 0.268 -0.950 3.590
8 67.550 0.418 -0.956 2.828
9 62.802 0.568 -0.964 2.062
10 62.108 0.637 -0.973 1.687
11 62.030 0.644 -0.979 1.619
12 62.003 0.647 -0.982 1.584
13 61.996 0.651 -0.985 1.549
14 61.996 0.651 -0.986 1.539
Unable to reduce sum of squares any further
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.6513 0.1434 4.54
MA 1 -0.9857 0.0516 -19.11
Constant 1.5385 0.4894 3.14
Mean 4.412 1.403
Number of observations: 33
Residuals: SS = 61.8375 (backforecasts excluded)
MS = 2.0613 DF = 30
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36 48
Chi-Square 9.6(DF=10) 17.0(DF=22) * (DF= *) *
(DF= *)
Forecasts from period 33
95 Percent Limits
Period Forecast Lower Upper Actual
34 3.1362 0.3216 5.9507
35 3.5810 -1.8180 8.9801
36 3.8708 -2.3061 10.0476
37 4.0594 -2.4192 10.5380
38 4.1823 -2.4201 10.7848
1;ح هJ!
ذج ا1!'
أي ان ا
252
zt = 1.54 + 0.65zt −1 + at − 0.99at −1 , at ∼ N ( 0, 2.06 )
7 هt إ;(ر4!B و42 ا
!"رOCNا% رات ا
!"
) وإJ&
( )
φˆ1 = 0.6513, s.e. φˆ1 = 0.1434, t = 4.54
θˆ = −0.9857, s.e. (θˆ ) = 0.0516, t = −19.11
1 1
δˆ = 1.5385, s.e. (δˆ ) = 0.4894, t = 3.14
σˆ 2 = 2.0613, with d . f . = 30
.421'"& )
"!
اV!L ان5
:7Bا1(
اo%b نsا
ACF of Residuals for CPIChnge
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8
Lag
PACF of Residuals for CPIChnge
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8
Lag
253
4 P V2ز1C V(;C 7Bا1(
ل أن اC 7Bا1(
78
ا7CاA
اw. وا
;ا7CاA
اw.أ
!ط ا
;ا
:7Bا1(
ا4"(= o%b'
،46. &;اz `ء أي.
اري3;
ر) ا
!رج ا
Histogram of the Residuals
(response is CPIChnge)
6
Frequency
-3 -2 -1 0 1 2 3 4
Residual
.|K
"\ ا. ';& (و2
:7Bا1(
7"(6
;!ل ا59 اw6[&
إI''
Normal Probability Plot of the Residuals
(response is CPIChnge)
2
Residual
-1
-2
-3
-2 -1 0 1 2
Normal Score
254
.(2JC 4"(= 7Bا1(
ل ان ا1J
أنV6;
.4(J; !
) اJ
ات:('C 5 V& 4 ;!
7
;
ا
) ا
Time Series Plot for CPIChnge
(with forecasts and their 95% confidence limits)
10
CPIChnge
5 10 15 20 25 30
Time
:7
;
آ4 ;!
اAR(2) ذج1!
(6C ول%
'د
MTB > arima 2 0 0 c2
Type Coef StDev T
AR 1 1.1872 0.1625 7.31
AR 2 -0.4657 0.1624 -2.87
Constant 1.3270 0.2996 4.43
Mean 4.765 1.076
Number of observations: 33
Residuals: SS = 88.6206 (backforecasts excluded)
MS = 2.9540 DF = 30
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36 48
Chi-Square 19.8(DF=10) 25.4(DF=22) * (DF= *) *
(DF= *)
255
1;ح هJ!
ذج ا1!'
أي ان ا
zt = 1.33 + 1.187 zt −1 − 0.4657 zt −1 + at , at ∼ N ( 0, 2.95)
7 هt إ;(ر4!B و42 ا
!"رOCNا% رات ا
!"
) وإJ&
( )
φˆ1 = 1.1872, s.e. φˆ1 = 0.1625, t = 7.31
φˆ = −0.4657, s.e. (θˆ ) = 0.1624, t = −2.87
2 1
δˆ = 1.327, s.e. (δˆ ) = 0.2996, t = 4.43
σˆ 2 = 2.954, with d . f . = 30
;(ر9 إ
اI''
H 0 : φ2 = 0
H1 : φ2 ± 0
48 59ا
φˆ2 −0.4657
t0 = = = −2.8676
( )
s.e. φ2ˆ 0.1624
&. O
P-value ا
ـL1
MTB > cdf -2.8676;
SUBC> t 30.
Cumulative Distribution Function
Student's t distribution with 30 DF
x P( X <= x)
-2.8676 0.0037
φ2 = 0 \ انN
X أي0.05 & HB أ7 وه0.0037 ويC O
P-value أي ا
ـ
. AR(2) ذج1!'
\ اN
7
;
.و
:#
يL أ،ذج1!
H`N وإ;ر أ4J.
ا4 ;!
ا4('& (
!ذج اى6C ول5
. AIC ` ا
!"ر2 وا;[م ا4('!
;(رات ا9ا
256
-ل
?zء 7ذج :4
1ف
%ول '.ء
!1ذج
!; 4
)z(t
-2.5 -9.1 -19.2 -33.1 -52.8 -76.7 -103.2
-132.4 -165.9 -204.0 -246.0 -291.3 -339.9 -391.4
-446.4 -504.8 -566.7 -631.6 -698.2 -766.7 -836.1
-905.3 -975.1 -1044.2 -1111.3 -1177.4 -1242.9 -1307.3
-1371.8 -1436.5 -1500.0 -1561.5 -1620.8 -1679.0 -1736.9
-1794.5 -1851.1 -1906.9 -1960.6 -2010.0 -2055.4 -2097.5
-2136.4 -2173.3 -2210.8 -2250.5 -2290.6 -2328.5 -2363.4
-2398.7 -2437.6 -2482.0 -2533.0 -2589.5 -2651.9 -2721.4
-2797.6 -2880.4 -2968.6 -3060.8 -3156.1 -3253.4 -3353.3
-3456.9 -3564.3 -3675.0 -3788.7 -3906.6 -4028.2 -4153.0
-4281.7 -4412.0 -4542.3 -4673.3 -4805.0 -4937.2 -5068.7
-5197.8 -5323.6 -5444.4 -5558.0 -5663.8 -5760.6 -5848.8
-5931.6 -6011.8 -6090.4 -6167.9 -6244.3 -6317.8 -6387.3
-6453.5 -6518.9 -6584.6 -6649.0 -6711.8 -6773.3 -6834.5
-6896.0 -6957.1 -7015.6 -7073.1 -7131.7 -7193.3 -7259.9
-7332.9 -7411.2 -7495.3 -7586.4 -7683.2 -7784.7 -7891.4
-8004.6 -8124.0 -8249.2 -8379.1 -8512.3 -8649.8 -8791.2
-8936.4 -9086.3 -9239.6 -9394.6 -9552.0 -9713.3 -9878.4
-10047.2 -10219.0 -10392.1 -10564.0 -10734.0 -10903.2 -11071.9
-11238.9 -11402.0 -11560.2 -11713.8 -11863.9 -12011.7 -12157.5
-12302.5 -12447.5 -12593.4 -12740.8 -12889.7 -13039.8 -13190.3
-13339.6 -13486.3 -13629.3 -13769.9 -13910.1 -14051.7 -14196.3
-14345.9 -14499.7 -14657.5 -14819.8 -14986.4 -15158.3 -15335.9
-15518.4 -15705.0 -15895.8 -16091.5 -16290.9 -16492.7 -16697.1
-16904.2 -17113.5 -17324.4 -17535.7 -17745.2 -17951.5 -18155.6
-18360.1 -18564.2 -18766.2 -18965.1 -19161.7 -19356.3 -19547.0
-19733.5 -19919.5 -20107.3 -20294.5 -20478.0 -20655.7 -20827.7
-20993.7 -21154.4 -21310.9 -21463.7 -21614.0 -21762.1 -21908.3
-22053.2 -22195.6 -22335.1 -22474.4
و
Oا
) ا
&' 7ا
;
:7
257
O r ig in a l T im e S e r ie s
-1 0 0 0 0
z(t)
-2 0 0 0 0
In d e x 50 100 150 200
:7 ه48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6.ا
;ا
Autocorrelation Function for z(t)
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
10 20 30 40 50
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 0.9813.93 196.91 13 0.80 2.492165.64 25 0.61 1.503459.42 37 0.43 0.984216.02 49 0.27 0.594589.24
2 0.97 8.00 388.73 14 0.78 2.372297.73 26 0.59 1.453540.78 38 0.42 0.944259.81 50 0.26 0.564607.24
3 0.95 6.15 575.44 15 0.76 2.252425.14 27 0.58 1.403618.50 39 0.41 0.914301.04
4 0.94 5.15 757.06 16 0.75 2.152547.94 28 0.56 1.353692.67 40 0.39 0.874339.79
5 0.92 4.50 933.59 17 0.73 2.062666.22 29 0.55 1.303763.38 41 0.38 0.844376.14
6 0.91 4.041105.06 18 0.72 1.972780.05 30 0.53 1.263830.74 42 0.36 0.814410.16
7 0.89 3.681271.47 19 0.70 1.892889.51 31 0.52 1.213894.84 43 0.35 0.774441.93
8 0.88 3.391432.85 20 0.68 1.822994.69 32 0.50 1.173955.76 44 0.34 0.744471.52
9 0.86 3.151589.23 21 0.67 1.753095.67 33 0.49 1.134013.61 45 0.32 0.714498.99
10 0.84 2.951740.66 22 0.65 1.683192.54 34 0.47 1.094068.47 46 0.31 0.684524.43
11 0.83 2.781887.17 23 0.64 1.623285.39 35 0.46 1.054120.44 47 0.30 0.654547.90
12 0.81 2.622028.81 24 0.62 1.563374.32 36 0.45 1.014169.59 48 0.28 0.624569.48
P artial A utocorrelation Function for z(t)
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
10 20 30 40 50
Lag PAC T Lag PAC T Lag PAC T Lag PAC T Lag PAC T
1 0.98 13.93 13 -0.01 -0.13 25 -0.01 -0.09 37 -0.01 -0.09 49 -0.01 -0.14
2 -0.01 -0.17 14 -0.01 -0.12 26 -0.01 -0.09 38 -0.01 -0.09 50 -0.01 -0.14
3 -0.01 -0.17 15 -0.01 -0.12 27 -0.01 -0.09 39 -0.01 -0.10
4 -0.01 -0.16 16 -0.01 -0.11 28 -0.01 -0.08 40 -0.01 -0.10
5 -0.01 -0.16 17 -0.01 -0.11 29 -0.01 -0.08 41 -0.01 -0.11
6 -0.01 -0.16 18 -0.01 -0.11 30 -0.01 -0.08 42 -0.01 -0.11
7 -0.01 -0.15 19 -0.01 -0.11 31 -0.01 -0.08 43 -0.01 -0.12
8 -0.01 -0.15 20 -0.01 -0.10 32 -0.01 -0.08 44 -0.01 -0.12
9 -0.01 -0.15 21 -0.01 -0.10 33 -0.01 -0.08 45 -0.01 -0.12
10 -0.01 -0.14 22 -0.01 -0.09 34 -0.01 -0.08 46 -0.01 -0.13
11 -0.01 -0.14 23 -0.01 -0.09 35 -0.01 -0.09 47 -0.01 -0.13
12 -0.01 -0.13 24 -0.01 -0.09 36 -0.01 -0.09 48 -0.01 -0.13
.w1;!
ا7N ةJ; & z zt 4 ;!
ا ان اL ^Pوا
O!
وwt = zt − zt −1
وXوق اb
اA{
258
F irs t D if f e re n c e s w (t)= z (t)-z (t-1 )
-1 0 0
w(t)
-2 0 0
In d e x 50 100 150 200
:7 ه4Bb!
ا4 ;!
48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6.ا
;ا
Autocorrelation Function for w(t)
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 15 25 35 45
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 0.9913.93 197.10 13 0.69 2.211983.03 25 0.39 1.082701.91 37 0.34 0.903047.50 49 0.31 0.763399.16
2 0.97 7.98 389.03 14 0.66 2.072076.38 26 0.38 1.032734.99 38 0.35 0.903077.32
3 0.95 6.12 574.74 15 0.63 1.932161.98 27 0.37 1.002766.16 39 0.35 0.903107.68
4 0.93 5.10 753.55 16 0.60 1.812240.33 28 0.36 0.962795.84 40 0.35 0.903138.42
5 0.91 4.43 924.81 17 0.57 1.702311.99 29 0.35 0.942824.46 41 0.35 0.903169.33
6 0.89 3.941087.93 18 0.54 1.602377.42 30 0.34 0.922852.35 42 0.35 0.893200.21
7 0.86 3.561242.44 19 0.52 1.502437.13 31 0.34 0.912879.79 43 0.35 0.883230.81
8 0.83 3.251388.10 20 0.49 1.412491.57 32 0.34 0.902907.03 44 0.34 0.873260.99
9 0.81 2.981524.83 21 0.47 1.332541.23 33 0.34 0.892934.32 45 0.34 0.853290.64
10 0.78 2.761652.58 22 0.45 1.262586.61 34 0.34 0.892961.85 46 0.33 0.843319.59
11 0.75 2.561771.42 23 0.43 1.192628.22 35 0.34 0.892989.82 47 0.33 0.823347.55
12 0.72 2.381881.49 24 0.41 1.132666.48 36 0.34 0.893018.34 48 0.32 0.793374.17
P a rtia l A u to c o rre la tio n F un c tio n fo r w (t)
Partial Autocorrelation
1 .0
0 .8
0 .6
0 .4
0 .2
0 .0
-0 .2
-0 .4
-0 .6
-0 .8
-1 .0
5 15 25 35 45
Lag P AC T Lag P AC T Lag P AC T Lag P AC T Lag P AC T
1 0 .9 9 1 3 .9 3 1 3 -0 .0 1 -0 .1 6 2 5 0 .0 4 0 .5 2 37 -0 .0 1 -0 .0 7 4 9 -0 .0 3 -0 .4 8
2 -0 .1 4 -1 .9 7 1 4 -0 .0 0 -0 .0 2 2 6 0 .0 4 0 .6 2 38 -0 .0 2 -0 .2 9
3 -0 .1 1 -1 .5 1 1 5 0 .0 1 0 .1 0 2 7 0 .0 3 0 .4 7 39 -0 .0 4 -0 .6 3
4 -0 .0 7 -0 .9 6 1 6 0 .0 2 0 .2 3 2 8 0 .0 4 0 .5 5 40 -0 .0 4 -0 .6 0
5 -0 .0 6 -0 .9 0 1 7 0 .0 2 0 .2 5 2 9 0 .0 4 0 .5 6 41 -0 .0 4 -0 .5 4
6 -0 .0 7 -0 .9 7 1 8 0 .0 1 0 .1 0 3 0 0 .0 2 0 .3 4 42 -0 .0 4 -0 .5 5
7 -0 .0 6 -0 .7 8 1 9 0 .0 1 0 .1 2 3 1 0 .0 2 0 .2 2 43 -0 .0 3 -0 .4 2
8 -0 .0 3 -0 .4 8 2 0 0 .0 1 0 .1 1 3 2 0 .0 1 0 .1 6 44 -0 .0 0 -0 .0 7
9 -0 .0 3 -0 .3 7 2 1 0 .0 1 0 .1 6 3 3 0 .0 1 0 .1 8 45 0 .0 1 0 .0 8
10 -0 .0 3 -0 .3 7 2 2 0 .0 3 0 .3 5 3 4 0 .0 1 0 .1 5 46 -0 .0 1 -0 .1 8
11 -0 .0 2 -0 .3 4 2 3 0 .0 2 0 .3 1 3 5 0 .0 0 0 .0 4 47 -0 .0 4 -0 .5 6
12 -0 .0 2 -0 .2 2 2 4 0 .0 2 0 .2 7 3 6 -0 .0 0 -0 .0 7 48 -0 .0 5 -0 .7 6
.w1;!
ا7N ةJ; & z الCX wt 4 ;!
ا ان اL ^Pوا
O!
( و4+ ا4 ;!
7
]
ق اb
ا اA ان ه5X) yt = wt − wt −1
وXوق اb
اA{
259
F ir s t D if f e r e n c e s y (t)= w (t)-w (t-1 )
1 0
y(t)
0
-5
In d e x 5 0 1 0 0 1 5 0 2 0 0
:7 ه4Bb!
ا4 ;!
@ اAO
48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6.ا
;ا
A utocorrelation Function for y(t)
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 15 25 35 45
Lag C orr T LB Q Lag C orr T LBQ Lag Corr T LBQ Lag C orr T LB Q Lag Corr T LBQ
1 0.9313.08 173.62 13 0.01 0.06 693.85 25 -0.64 -2.731124.21 37 -0.09 -0.301748.28 49 0.48 1.562017.77
2 0.80 6.83 303.36 14 -0.05 -0.23 694.31 26 -0.66 -2.721224.05 38 -0.01 -0.031748.29
3 0.70 4.89 401.81 15 -0.09 -0.46 696.12 27 -0.66 -2.621324.28 39 0.08 0.261749.70
4 0.62 3.89 479.62 16 -0.14 -0.70 700.34 28 -0.63 -2.441418.06 40 0.15 0.501755.14
5 0.56 3.26 542.93 17 -0.19 -0.97 708.51 29 -0.59 -2.221500.63 41 0.20 0.681765.20
6 0.50 2.77 593.77 18 -0.25 -1.24 722.16 30 -0.55 -2.011572.45 42 0.23 0.781778.55
7 0.43 2.32 632.35 19 -0.31 -1.53 743.30 31 -0.50 -1.801632.42 43 0.24 0.821793.44
8 0.36 1.90 659.84 20 -0.37 -1.83 774.40 32 -0.43 -1.531677.44 44 0.25 0.861810.08
9 0.29 1.50 677.66 21 -0.44 -2.13 818.29 33 -0.36 -1.251708.46 45 0.28 0.961831.09
10 0.22 1.11 687.71 22 -0.51 -2.40 876.73 34 -0.29 -1.001728.63 46 0.34 1.131860.92
11 0.15 0.75 692.41 23 -0.56 -2.57 948.00 35 -0.22 -0.761740.51 47 0.40 1.341903.36
12 0.08 0.41 693.82 24 -0.60 -2.671030.79 36 -0.16 -0.531746.42 48 0.45 1.491956.81
P artial A utocorrelation Function for y(t)
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 15 25 35 45
Lag PAC T Lag PAC T Lag PAC T Lag PAC T Lag PAC T
1 0.93 13.08 13 -0.05 -0.75 25 -0.03 -0.46 37 0.08 1.16 49 0.07 1.00
2 -0.46 -6.44 14 0.05 0.67 26 -0.03 -0.44 38 0.11 1.56
3 0.33 4.71 15 -0.07 -0.95 27 0.00 0.04 39 0.04 0.51
4 -0.14 -1.97 16 -0.09 -1.30 28 0.08 1.13 40 0.04 0.52
5 0.12 1.72 17 -0.04 -0.55 29 -0.01 -0.14 41 -0.05 -0.74
6 -0.13 -1.78 18 -0.08 -1.11 30 -0.04 -0.56 42 -0.05 -0.66
7 0.00 0.01 19 -0.12 -1.71 31 0.15 2.17 43 -0.14 -1.95
8 -0.06 -0.81 20 -0.11 -1.50 32 0.06 0.88 44 -0.01 -0.10
9 -0.08 -1.11 21 -0.15 -2.12 33 -0.00 -0.05 45 0.07 0.92
10 -0.04 -0.55 22 -0.07 -0.93 34 0.04 0.49 46 0.06 0.79
11 -0.04 -0.58 23 -0.06 -0.87 35 0.03 0.46 47 -0.01 -0.17
12 -0.09 -1.21 24 -0.15 -2.12 36 0.04 0.50 48 -0.09 -1.24
ةJ; & S%(+ اO ا48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6. و ا
;ا4 ;!
اH3 & 5
. d=2 اي انw1;!
ا7N
260
!& اولe[;
;[& & اC O
ى ا48
ا4CاA
ت ا6. وا
;ا4CاA
ت ا6.& ا
!ط ا
;ا
&. ذج1!'
ا اA( ه6
ف1 وzt 4+ ا4 ;!
ARIMA(1,2,1) ذج1!
^2
MTB > ARIMA 1 2 1 'z(t)' 'RESI2' 'FITS2';
SUBC> NoConstant;
SUBC> Forecast 10 c4 c5 c6;
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot.
ARIMA Model
ARIMA model for z(t)
Estimates at each iteration
Iteration SSE Parameters
0 2462.77 0.100 0.100
1 1345.58 0.250 -0.050
2 1170.63 0.203 -0.200
3 984.83 0.182 -0.350
4 782.47 0.200 -0.500
5 560.15 0.278 -0.650
6 363.93 0.428 -0.765
7 259.20 0.578 -0.814
8 202.76 0.728 -0.842
9 185.51 0.861 -0.859
10 185.36 0.873 -0.860
11 185.36 0.875 -0.860
12 185.36 0.875 -0.860
Relative change in each estimate less than 0.0010
Final Estimates of Parameters
Type Coef StDev T
261
AR 1 0.8749 0.0353 24.75
MA 1 -0.8599 0.0357 -24.12
Differencing: 2 regular differences
Number of observations: Original series 200, after
differencing 198
Residuals: SS = 183.717 (backforecasts excluded)
MS = 0.937 DF = 196
Modified Box-Pierce (Ljung-Box) Chi-Square statistic
Lag 12 24 36 48
Chi-Square 3.9(DF=10) 13.0(DF=22) 33.1(DF=34)
46.0(DF=46)
Forecasts from period 200
95 Percent Limits
Period Forecast Lower Upper Actual
201 -22615.2 -22617.1 -22613.3
202 -22757.2 -22764.6 -22749.9
203 -22900.4 -22917.2 -22883.5
204 -23044.5 -23075.3 -23013.7
205 -23189.5 -23238.8 -23140.1
206 -23335.2 -23407.8 -23262.6
207 -23481.5 -23582.1 -23380.9
208 -23628.4 -23761.8 -23495.0
209 -23775.8 -23946.7 -23605.0
210 -23923.7 -24136.6 -23710.7
1;ح هJ!
ذج ا1!'
ا
zt = 0.875t −1 z + at − 0.859at −1 , at ∼ N ( 0,0.937 )
7 هt إ;(ر4!B و42 ا
!"رOCNا% رات ا
!"
) وإJ&و
( )
φˆ1 = 0.8749, s.e. φˆ1 = 0.0353, t = 24.75
θˆ1 = −0.8599, s.e. (θˆ ) = 0.0357,
1 t = −24.12
σˆ 2 = 0.937, with d . f . = 196
262
.421'"& )
"!
ان ا5
:7Bا1(
اo%b
نsا
ACF of Residuals for z(t)
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 10 15 20 25 30 35 40 45
Lag
PACF of Residuals for z(t)
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
5 10 15 20 25 30 35 40 45
Lag
4 P V2ز1C V(;C 7Bا1(
ل أن اC 7Bا1(
78
ا7CاA
اw. وا
;ا7CاA
اw.أ !ط ا
;ا
:7Bا1(
ا4"(= o%b'
،46. &;اz `ء أي.
اري3;
ر) ا
!رج ا
263
Histogram of the Residuals
(response is z(t))
20
Frequency
10
-3 -2 -1 0 1 2
Residual
.|K
"\ ا. ';& (و2
:7Bا1(
7"(6
;!ل ا59 اw6[&
إI''
Normal Probability Plot of the Residuals
(response is z(t))
1
Residual
-1
-2
-3
-3 -2 -1 0 1 2 3
Normal Score
.(2JC 4"(= 7Bا1(
ل ان ا1J أنV6;
.:('C ;اتN 95% V& 4(J; !
) اJ
ات:('C 10
ـ7
;
ا
) ا
Forecast of 20 Future values with 95% limits
-22500
-23000
z(t)
-23500
-24000
0 1 2 3 4 5 6 7 8 9 10
Time
264
ﻤﻠﺤﻕ )(1
أ v$ ~0ا?jرات ا $و v$ا$njت اL /
265
ﺒﺴﻡ ﺍﷲ ﺍﻝﺭﺤﻤﻥ ﺍﻝﺭﺤﻴﻡ
ﻗﺴﻡ ﺍﻻﺤﺼﺎﺀ ﻭﺒﺤﻭﺙ ﺍﻝﻌﻤﻠﻴﺎﺕ
ﻜﻠﻴﺔ ﺍﻝﻌﻠﻭﻡ ﺠﺎﻤﻌﺔ ﺍﻝﻤﻠﻙ ﺴﻌﻭﺩ
ﺍﻻﺨﺘﺒﺎﺭ ﺍﻝﻨﻬﺎﺌﻰ ﻝﻠﻔﺼل ﺍﻻﻭل 1420/1419ﻫـ
ﺍﻝﻤﺎﺩﺓ 221ﺒﺤﺙ ﻁﺭﻕ ﺍﻝﺘﻨﺒﺅ ﺍﻻﺤﺼﺎﺌﻰ
ﺍﺠﺏ ﻋﻠﻰ ﺠﻤﻴﻊ ﺍﻻﺴﺌﻠﺔ ﺍﻝﺘﺎﻝﻴﺔ:
ﺍﻝﺴﺅﺍل ﺍﻻﻭل:
ﺍﻝﺒﻴﺎﻨﺎﺕ ﺍﻝﺘﺎﻝﻴﺔ ﺘﻤﺜل ﻋﺩﺩ ﺍﻝﺴﻴﺎﺭﺍﺕ ﺍﻝﻤﺒﺎﻋﺔ ﺍﺴﺒﻭﻋﻴﺎ ﻝﺩﻱ ﻤﻭﺯﻉ ﻤﺎ
10 9 8 7 6 5 4 3 2 1 ﺍﻻﺴﺒﻭﻉ
85 77 80 71 78 69 83 79 75 75 ﺍﻝﻌﺩﺩ
ﺍﻭﺠﺩ ﺘﻨﺒﺅﺍﺕ ﻝﻌﺩﺩ ﺍﻝﺴﻴﺎﺭﺍﺕ ﺍﻝﺘﻰ ﺴﺘﺒﺎﻉ ﻓﻰ ﺍﻻﺴﺒﻭﻋﻴﻥ ﺍﻝﺘﺎﻝﻴﻴﻥ ﻭﺍﻭﺠﺩ ﻓﺘﺭﺍﺕ ﺘﻨﺒﺅ 95%ﻝﻬﺫﻩ
ﺍﻝﺘﻨﺒﺅﺍﺕ ﻜﻠﻤﺎ ﺍﻤﻜﻥ ﺫﻝﻙ ﺒﺎﺴﺘﺨﺩﺍﻡ:
ﺍ( ﻨﻤﻭﺫﺝ ﺍﻨﺤﺩﺍﺭ ﺨﻁﻰ ﻝﻠﻌﺩﺩ ﺍﻝﻤﺒﺎﻉ ﻤﻊ ﺍﻝﺯﻤﻥ ﺒﺎﻻﺴﺎﺒﻴﻊ.
ﺏ( ﺍﻝﺘﻤﻬﻴﺩ ﺒﻭﺍﺴﻁﺔ ﻤﺘﻭﺴﻁ ﻤﺘﺤﺭﻙ ﻤﻥ ﺍﻝﺩﺭﺠﺔ ﺍﻝﺜﺎﻝﺜﺔ.
ﺝ( ﺍﻝﺘﻤﻬﻴﺩ ﺍﻻﺴﻰ ﺍﻝﺒﺴﻴﻁ ﻤﺴﺘﺨﺩﻤﺎ . α = 0.3
ﺍﻝﺴﺅﺍل ﺍﻝﺜﺎﻨﻲ:
ﻝﻠﻨﻤﻭﺫﺝ (1 − φ1B − φ 2 B2 ) zt = δ + (1 − θB) atﺤﻴﺙ ) at ~ WN (0,σ 2ﻭ φ1 , φ2 ,δ ,θﻫﻰ ﻤﻌﺎﻝﻡ
ﺍﻝﻨﻤﻭﺫﺝ ﻭ Bﻫﻭ ﻋﺎﻤل ﺍﻻﺯﺍﺤﺔ ﺍﻝﺨﻠﻔﻰ ﺍﻭﺠﺩ:
ﺍ( E ( zt ) ∀t
ﺏ( ﺩﺍﻝﺔ ﺍﻝﺘﺭﺍﺒﻁ ﺍﻝﺫﺍﺘﻰ ρk ∀k ≥ 0
ﺝ( ﺩﺍﻝﺔ ﺍﻝﺘﺭﺍﺒﻁ ﺍﻝﺫﺍﺘﻰ ﺍﻝﺠﺯﺌﻰ φ kk ∀k ≥ 0
ﺩ( ﺩﺍﻝﺔ ﺍﻻﻭﺯﺍﻥ ψ j ∀j ≥ 0
ﺍﻝﺴﺅﺍل ﺍﻝﺜﺎﻝﺙ:
266
zt = 38.5 + 12ﺤﻴﺙ ) at ~ WN (0,4ﻭﺍﺫﺍ ﻜﺎﻨﺕ t = 10ﻭ
ﻝﻠﻨﻤﻭﺫﺝ . zt −1 − 0.7 zt − 2 + at − 0.4at −1
z9 = 77ﻭ z10 = 85ﻭ a10 = −1.6ﺍﻭﺠﺩ:
ﺍ( ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺍﻝﺘﻰ ﻝﻬﺎ ﺍﺩﻨﻰ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﺨﻁﺎﺀ
ﺏ( ﺘﺒﺎﻴﻥ ﺩﺍﻝﺔ ﺍﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﺤﺘﻰ ﺯﻤﻥ ﺍﻝﺘﻘﺩﻡ ℓ = 3
ﺝ( ﺘﻨﺒﺅﺍﺕ ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ z11ﻭ z12
ﺩ( ﻓﺘﺭﺍﺕ ﺘﻨﺒﺅ 95%ﻝﻠﺘﻨﺒﺅﺍﺕ ﺍﻝﺴﺎﺒﻘﺔ
ﻫـ( ﺍﺫﺍ ﻋﻠﻤﺕ ﺍﻥ z11 = 81ﻓﺠﺩﺩ ﺍﻝﺘﻨﺒﺅ ﻝﻠﻘﻴﻤﻪ z12ﻭﻝﻔﺘﺭﺓ ﺘﻨﺒﺅﻫﺎ.
267
ﺒﺴﻡ ﺍﷲ ﺍﻝﺭﺤﻤﻥ ﺍﻝﺭﺤﻴﻡ
ﻗﺴﻡ ﺍﻻﺤﺼﺎﺀ ﻭﺒﺤﻭﺙ ﺍﻝﻌﻤﻠﻴﺎﺕ
ﻜﻠﻴﺔ ﺍﻝﻌﻠﻭﻡ ﺠﺎﻤﻌﺔ ﺍﻝﻤﻠﻙ ﺴﻌﻭﺩ
ﺍﻻﺨﺘﺒﺎﺭ ﺍﻝﻨﻬﺎﺌﻰ ﻝﻠﻔﺼل ﺍﻻﻭل 1420/1419ﻫـ
ﺍﻝﻤﺎﺩﺓ 221ﺒﺤﺙ ﻁﺭﻕ ﺍﻝﺘﻨﺒﺅ ﺍﻻﺤﺼﺎﺌﻰ
ﺍﺠﺏ ﻋﻠﻰ ﺠﻤﻴﻊ ﺍﻻﺴﺌﻠﺔ ﺍﻝﺘﺎﻝﻴﺔ:
ﺍﻝﺴﺅﺍل ﺍﻻﻭل:
ﺍﻝﺒﻴﺎﻨﺎﺕ ﺍﻝﺘﺎﻝﻴﺔ ﺘﻤﺜل ﻋﺩﺩ ﺍﻝﺴﻴﺎﺭﺍﺕ ﺍﻝﻤﺒﺎﻋﺔ ﺍﺴﺒﻭﻋﻴﺎ ﻝﺩﻱ ﻤﻭﺯﻉ ﻤﺎ
Week 1 2 3 4 5 6 7 8 9 10
No. of cars 75 75 79 83 69 78 71 80 77 85
ﺍﻭﺠﺩ ﺘﻨﺒﺅﺍﺕ ﻝﻌﺩﺩ ﺍﻝﺴﻴﺎﺭﺍﺕ ﺍﻝﺘﻰ ﺴﺘﺒﺎﻉ ﻓﻰ ﺍﻻﺴﺒﻭﻋﻴﻥ ﺍﻝﺘﺎﻝﻴﻴﻥ ﻭﺍﻭﺠﺩ ﻓﺘﺭﺍﺕ ﺘﻨﺒﺅ 95%ﻝﻬﺫﻩ
ﺍﻝﺘﻨﺒﺅﺍﺕ ﻜﻠﻤﺎ ﺍﻤﻜﻥ ﺫﻝﻙ ﺒﺎﺴﺘﺨﺩﺍﻡ:
ﺍ( ﻨﻤﻭﺫﺝ ﺍﻨﺤﺩﺍﺭ ﺨﻁﻰ ﻝﻠﻌﺩﺩ ﺍﻝﻤﺒﺎﻉ ﻤﻊ ﺍﻝﺯﻤﻥ ﺒﺎﻻﺴﺎﺒﻴﻊ.
ﺏ( ﺍﻝﺘﻤﻬﻴﺩ ﺒﻭﺍﺴﻁﺔ ﻤﺘﻭﺴﻁ ﻤﺘﺤﺭﻙ ﻤﻥ ﺍﻝﺩﺭﺠﺔ ﺍﻝﺜﺎﻝﺜﺔ.
ﺝ( ﺍﻝﺘﻤﻬﻴﺩ ﺍﻻﺴﻰ ﺍﻝﺒﺴﻴﻁ ﻤﺴﺘﺨﺩﻤﺎ . α = 0.3
ﺍﻝﺴﺅﺍل ﺍﻝﺜﺎﻨﻲ:
ﻝﻠﻨﻤﻭﺫﺝ (1 − φ1B − φ 2 B2 ) zt = δ + (1 − θB) atﺤﻴﺙ ) at ~ WN (0,σ 2ﻭ φ1 , φ2 ,δ ,θﻫﻰ ﻤﻌﺎﻝﻡ
ﺍﻝﻨﻤﻭﺫﺝ ﻭ Bﻫﻭ ﻋﺎﻤل ﺍﻻﺯﺍﺤﺔ ﺍﻝﺨﻠﻔﻰ ﺍﻭﺠﺩ:
ﺍ( E ( zt ) ∀t
ﺏ( ﺩﺍﻝﺔ ﺍﻝﺘﺭﺍﺒﻁ ﺍﻝﺫﺍﺘﻰ ρk ∀k ≥ 0
ﺝ( ﺩﺍﻝﺔ ﺍﻝﺘﺭﺍﺒﻁ ﺍﻝﺫﺍﺘﻰ ﺍﻝﺠﺯﺌﻰ φ kk ∀k ≥ 0
ﺩ( ﺩﺍﻝﺔ ﺍﻻﻭﺯﺍﻥ ψ j ∀j ≥ 0
ﺍﻝﺴﺅﺍل ﺍﻝﺜﺎﻝﺙ:
268
zt = 38.5 + 12ﺤﻴﺙ ) at ~ WN (0,4ﻭﺍﺫﺍ ﻜﺎﻨﺕ t = 10ﻭ
ﻝﻠﻨﻤﻭﺫﺝ . zt −1 − 0.7 zt − 2 + at − 0.4at −1
z9 = 77ﻭ z10 = 85ﻭ a10 = −1.6ﺍﻭﺠﺩ:
ﺍ( ﺩﺍﻝﺔ ﺍﻝﺘﻨﺒﺅ ﺍﻝﺘﻰ ﻝﻬﺎ ﺍﺩﻨﻰ ﻤﺘﻭﺴﻁ ﻤﺭﺒﻊ ﺍﺨﻁﺎﺀ
ﺏ( ﺘﺒﺎﻴﻥ ﺩﺍﻝﺔ ﺍﺨﻁﺎﺀ ﺍﻝﺘﻨﺒﺅ ﺤﺘﻰ ﺯﻤﻥ ﺍﻝﺘﻘﺩﻡ ℓ = 3
ﺝ( ﺘﻨﺒﺅﺍﺕ ﻝﻠﻘﻴﻡ ﺍﻝﻤﺴﺘﻘﺒﻠﻴﺔ z11ﻭ z12
ﺩ( ﻓﺘﺭﺍﺕ ﺘﻨﺒﺅ 95%ﻝﻠﺘﻨﺒﺅﺍﺕ ﺍﻝﺴﺎﺒﻘﺔ
ﻫـ( ﺍﺫﺍ ﻋﻠﻤﺕ ﺍﻥ z11 = 81ﻓﺠﺩﺩ ﺍﻝﺘﻨﺒﺅ ﻝﻠﻘﻴﻤﻪ z12ﻭﻝﻔﺘﺭﺓ ﺍﻝﺘﻨﺒﺅ
269
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
ﻗﺴﻡ ﺍﻹﺤﺼﺎﺀ ﻭﺒﺤﻭﺙ ﺍﻝﻌﻤﻠﻴﺎﺕ
ا
!دة = :ق ا
;'( :اQ%. 221 78 5X
ﺍﻻﺨﺘﺒﺎﺭ ﺍﻻﻭل ﻝﻸﻋﻤﺎل ﺍﻝﻔﺼﻠﻴﺔ
ا
H bاول 1421-1420هـ
ا
& ; :
أ V!L 7 WLا 4ا
;
:4
ا
:ال اول 4 ;!
:ز&'{zt } 4
-1أذآ وط اJ;9ار
-2ف دا
4ا
;ا w.ا
Aا7C
ا
:ال ا
]
4 ;!
:7ا
;
4
t 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
zt 53 43 66 48 52 42 44 56 44 58 41 54 51 56 38
w (= -1إ %ار ( t , zt ) .و& )Uأوz16 , z17 L
%;& w1;& (= -2ك & ا
ر 4Lا
]
] 4و& )Uأوz16 , z17 L
O!C (= -3ا α = 0.5 w . 7و& )Uأوz16 , z17 L
-4إذا آ z16 = 56 Sو {N z17 = 49ي & ا
6ق ا
4J.أآ] د 4Bوذ
[;F. fام &"ر
W'& {6
270
:Q
]
ال ا:
ا
zt = 20 − 0.9 zt −1 + at , at ~ WN ( 0, 4 ) ذج1!'
؟J; & ذج1!'
اH ه-1
µ L أو-2
k = 0,1,2,...,5 )J
φkk وρ k L أو-3
271
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
ﻗﺴﻡ ﺍﻹﺤﺼﺎﺀ ﻭﺒﺤﻭﺙ ﺍﻝﻌﻤﻠﻴﺎﺕ
ﺍﻝﻤﺎﺩﺓ :ﻁﺭﻕ ﺍﻝﺘﻨﺒﺅ ﺍﻻﺤﺼﺎﺌﻲ 221ﺒﺤﺙ
ﺍﻻﺨﺘﺒﺎﺭ ﺍﻝﻨﻬﺎﺌﻲ ﻝﻠﻔﺼل ﺍﻝﺜﺎﻨﻲ 1421-1420ﻫـ
ﺍﻝﺯﻤﻥ :ﺃﺭﺒﻊ ﺴﺎﻋﺎﺕ
أ V!L 7 WLا 4ا
;
:4
ا
:ال اول( 4& 20 ) :
أآ! Hا
bاzت ا
;
:4
(
γ k = Cov Z t , Zودا
4
!; 4ز&' {Z } 4دا
4ا
; 2Yا
Aا) 4B "
. 6"C 7C
t (1
ρkو = ρ0و 1 ρ k = γ kو
Oا
[1اص ا
;ا w.ا
Aا4B "
. 7C
& 5 ) ρ k = ρت (
ا
" 4B 4 ;!
(2ز&'K& 4هة z1 , z2 ,⋯ , znدا
4ا
;ا w.ا
AاJC 4'"
7Cر &
n
) rk = ∑ ( zt − z )( zt +k − zو ا%
9اف ا
!"ري ∑( z t ⋯− z ) , k = 0,1,
i =1 i =1
s.e ( rk ) ≃ 1آ! ان دا
4ا
;ا w.ا
Aا7C
ا
4ا
;ا w.ا
AاJC 4'"
7Cر & ا
" 4B
k
rk +1 − ∑ rkj r
j =1
= rk +1,k +1و ا
" 4Bا
! ة k
ا
" 4B ا
JC 4'"
78ر &
1− ∑r rj
j =1
& 8 ) ±ت ( rk +1, j = rkj − rk +1,k +1rk ,و
5 Oود 4L5 , j = 1,...,
n
272
4B "
. 6"2 AR ( ) ذج1!
(3
H3K
ا7 W;32 يA
و ا1 − φ1 ( − φ2
2
) Z = a , a ∼ WN (
t t t , )
( &ت7 ) Z t = δ + φ1 + φ2 + at
( 4& 40 ) :7
]
ال ا:
ا
( !
أ & ا
رB )إ:تX2
ف اX. & 4آK
4&1
) ا
!("ت ا5 7 ه4
;
ا
(
ت ا
29.3 20.0 25.8 29.0 31.0 32.7 33.6
27.5 26.8 30.6
28.9 28.5 28.2 26.1 27.8 27.6 29.9
28.2 26.7 30.0
30.8 30.5 36.6 31.4 30.8 33.2 30.2
27.1 33.7 36.6
29.0 28.1 30.3 29.4 33.6 30.3 20.1
17.5 23.7 24.2
32.4 32.4 29.4 23.5 23.6 28.1 29.9
30.6 32.3 31.6
28.0 24.1 29.2 34.3 26.4 28.8 21.3
21.7 21.5 24.7
33.6 36.5 35.7 33.7 29.3 25.1 27.2
30.6 29.1 28.5
32.0 31.9 31.7 29.0 31.9 24.3 22.7
26.6 28.9 28.3
28.2 28.6 30.7 30.6 20.8 16.6 25.2
31.8 32.5 30.3
26.1 19.0 24.3 31.5 32.0 31.7 29.1
23.2
48 & ذج1!'
ا7 "فC Minitab 7N %pacf و%acf 2&;[ام اF. (1
( &ت7 ) ;("* ا
( تC يA
واARIMA ( p, d , q ) ا
'!ذج
273
"ت ا.!
ا4J2= ا
"وم و4J26. * ذج ا
!;"ف1!'
!"
) ا4
رات أوJ& L( أو2
( &ت9 ) 4=K
ا
7 ذج1!'
اW; أو ا
"وم أآ4=K
"ت ا
ا.!
ا4J26. )
"!
رات اJ& ;[امF. (3
Q5 Z t = δ + φ1Z t −1 + ⋯ + φ p Z t − p + at − θ1at −1 − ⋯ − θ q at −q H3K
ا
( &ت4 ) at ∼ WN ( 0, σ 2 )
أنS! وإذاZ100 وZ 99 4(J; !
) اJ
ات:('C L) أوU &ذج و1!'
:(';
ا4
داL( أو4
( &ت10) Z100 وZ 99 4(J; !
) اJ
95% :('C ;اتN LوN σ 2 = 10.83
HON z100 = 32.4 S وإذا آZ100 4(J; !
ا4!J
:(';
د اN z99 = 26.7 أنS! ( إذا5
( &ت5 ) ؟2 ;
" ا. امH(B H`N أ:(';
ا
وforecast 5 C2 C3 C4 4b
واوا& اarima p d q C1 &;[ام اF. (6
f 8; V& رنB وJ%C gfit وgnormal وghist وgpacf وgacf وgseries
( &ت5 ) .4J.
ات اJb
ا7N
274
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
) Bا 59ء و1%.ث ا
"!ت
إ;(ر ا
H bاول 1421هـ1422/هـ
!دة = ) Q%. 221ق ا
;'( :ا( 78 59
ﺍﻝﺯﻤﻥ 3ﺴﺎﻋﺎﺕ
أ V!L WLا 4ا
;
:4
ا
:ال اول:
-1ف ا
;
;F. 7ر :ا
` 4ا
(`ء ،ا
;ا w.ا
Aا ،7Cا
;! ،OاJ;9ار
-2اآ; Wا
!"دXت ا
!"!'
4Nذج ا
;
AR(2), MA(1), ARMA(1,2) :4
ج(
'!1ذج ) AR(2أي & ا
!"
) ا
;
J%C 4اJ;9ار:
1) φ1 = 1.2, φ2 = −0.8
2) φ1 = −1.2, φ2 = −0.8
3) φ1 = 0.8, φ2 = −0.8
4) φ1 = −0.8, φ2 = 1.2
ا
:ال ا
]
:7
ﺍﻝﺒﻴﺎﻨﺎﺕ ﺍﻝﺘﺎﻝﻴﺔ ﺘﻤﺜل ﻤﺒﻴﻌﺎﺕ ﺃﺠﻬﺯﺓ ﺍﻝﺤﺎﺴﺏ ﻓﻲ ﺃﺤﺩ ﺍﻝﺸﺭﻜﺎﺕ ﺸﻬﺭﻴﺎ ) ﺇﻗﺭﺃ ﻤﻥ ﺍﻝﻴﺴﺎﺭ ﻝﻠﻴﻤﻴﻥ
ﺴﻁﺭﺍ ﺒﺴﻁﺭ(
26 21 16 28 27 19
21 26 25 16 20 25
25 21 23 26 21 23
17 25
-1أد Hا
(
ت 7Nور Minitab _&
(
H! 4Bو أر! Oآ!; 4ز&'.4
(= -2ا
(
ت ا
'!ذج ا
;
:4
i) Linear Trend Model
ii) Simple Moving Average Model of order 3
iii) Simple Exponential Smoothing Model with α =0.3
275
ج( =( ا
(
ت
!1ذج & ARMA( p, q) 48وذ
";
. fف p, qا
!'(4
و& B )Uر ا
!"
)
'!1ذج ا
!;Jح.
د( أي
!1ذج & ا
'!ذج ا
7N ) 4J.ا
CJbب و ج ( e 2ا
!Kهات H3K.أH`N؟
هـ( [;F.ام ا
'!1ذج ا H`Nو
:('Cات
2OKا
;
;b.ات 95% :('C
276
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
ﻗﺴﻢ ﺍﻹﺣﺼﺎﺀ ﻭﺑﺤﻮﺙ ﺍﻟﻌﻤﻠﻴﺎﺕ
ﺍﻟﻤﺎﺩﺓ :ﻃﺮﻕ ﺍﻟﺘﻨﺒﺆ ﺍﻻﺣﺼﺎﺋﻲ 221ﺑﺤﺚ
ﺍﻻﺧﺘﺒﺎﺭ ﺍﻻﻭﻝ ﻟﻸﻋﻤﺎﻝ ﺍﻟﻔﺼﻠﻴﺔ
ﺍﻟﻔﺼﻞ ﺍﻟﺜﺎﻧﻲ 1422/1421ﻫـ
ا
& ; :
أ V!L 7 WLا 4ا
;
:4
ا
:ال اول1 :هت ا
!; 4ا
&' 4ا
;
) :4إBا & ا
ر
! 6ا (6 .
47 64 23 71 38
64 55
41 59 48 71 35
57 40
58 44 80 55 37
74 51
57 50 60 45 57
50 45
25 59 50 71 56
74 50
58 45 54 36 54
48 55
45 57 50 62 44
64 43
52 38 59 55 41
53 49
34 35 54 45 68
38 50
277
60 39 59 40 57
54 23
:7
;
اWL وأMinitab _&
. 7N Worksheet H! 4b%+ 7N هاتK!
اHأد
(ة5 وا4& ) ة؟
!ذا؟J; & 4 ;!
(و اC H هTsplot 46ا1. هاتK!
اo%N)أ( إ
4& ) ؟ و
!ذا؟4 ;!
@ اAO
Trend Analysis
اف9 اH%C 4J2=
;[مH)ب( ه
(ة5وا
4& ) ؟ و
!ذا؟4 ;!
@ اAO
Decomposition Method f3b;
ا4J2=
;[مH)ج( ه
(ة5وا
@AO
Moving Average Smoothing ك%;!
اw1;!
. O!;
ا4J2=
;[مH)د( ه
(ة5 وا4& ) ؟ و
!ذا؟4 ;!
ا
@AO
Simple Exponential Smoothing w (
ا7X اO!;
ا4J2=
;[مH)هـ( ه
(ة5 وا4& ) ؟ و
!ذا؟4 ;!
ا
@AO
Double Exponential Smoothing 78']
ا7X اO!;
ا4J2=
;[مH)و( ه
(ة5 وا4& ) ؟ و
!ذا؟4 ;!
ا
(ة5 وا4& ) ؟ و
!ذا؟4 ;!
@ اAO
Winters’ Method و
;ز4J2=
;[مH)ز( ه
4'&
ا4 ;!
اH%;
W
ا7 هwJN ة5 وا4J2= L1C .
ال ا:
& ا:7
]
ال ا:
ا
. 95% :('C ;اتN V& 4(J; & )B 4 ![
ات:('C L أو4('!
ا4J26
;[ام اF. !هةK!
ا
( &ت8)
278
إ 4.Lا
:ال اول:
)أ (
80
70
60
Sales
50
40
30
20
Index 10 20 30 40 50 60 70
ا
!; (C 4و & ;Jة Q5ا
15 Y;C Oل & ;1ى S.U
)ب( 3!2إ;[ام = H%C 4J2ا
9ف و
4('& z O'3ه'
"م و1Lد إ* ;C "& @ C
* ا
!; 4
)ج( 3!2إ;[ام = 4J2ا
; f3bو
W'C O'3أآ] ا
!; ت ا
;& ON 7آ(ت إ
اف و
&4!1
)د( W'2ا
;! w1;!
. Oا
!;%ك هAا ا
'1ع & ا
!; ت ا
&' 4أآ] & & @z
ا
6ق
* آ!
ى ه @Aا
!; ;%C 4ج إ
w . O!Cإذ
~ O.أي إ
اف أو &4!1
)هـ( ا
;! Oا 7Xا
( H]!
Vb'2 wه @Aا
!; ت أ `2و 4+إذا آ
1%C Sي إ
اN
w .
)و( ا
;! Oا 7Xا
]' b2 78أآ] 4
5 7Nا
!; ت ا
;1%C 7ي إ
ا 76 z Nوا
Aي
(2Xو & Cف ا
!; 4ا
!Kهة
)ز( = 4J2و
;ز ;!
Vb'Cت ا
! 4!1ا
;1%C 7ي &آ( 4إ
اف 76 z
إ 4.Lا
:ال ا
]
:7
(2و أن = 4J2ا
;! w1;!
. Oا
!;%ك ه 7اآ] &'( H%;
4ا
!; 4ا
!Kهة '
.ب
ة &;61ت &;%آ O!;
4ا
!; 4وإ 2د :('Cات
.O
) '
(1ب &;%;& w1ك & ا
ر 4Lا
]
!&) 4آ(:
;MTB > %MA 'Sales' 2
279
SUBC> Center;
SUBC> Forecasts 5.
Executing from file: G:\MTBWIN\MACROS\MA.MAC
Macro is running ... please wait
Moving average
Data Sales
Length 70.0000
NMissing 0
Moving Average
Length: 2
Accuracy Measures
MAPE: 21.843
MAD: 10.045
MSD: 154.429
Row Period Forecast Lower Upper
1 71 47 22.6432 71.3568
2 72 47 22.6432 71.3568
3 73 47 22.6432 71.3568
4 74 47 22.6432 71.3568
5 75 47 22.6432 71.3568
280
Moving Average
Actual
80
Predicted
70 Forecast
Actual
Predicted
60
Forecast
Sales
50
Moving Average
40 Length: 2
30 MAPE: 21.843
MAD: 10.045
20 MSD: 154.429
0 10 20 30 40 50 60 70
Time
:4]
]
ا4Lك & ا
ر%;& w1;& (
' ب2)
MTB > %MA 'Sales' 3;
SUBC> Forecasts 5.
Executing from file: G:\MTBWIN\MACROS\MA.MAC
Macro is running ... please wait
Moving average
Data Sales
Length 70.0000
NMissing 0
Moving Average
Length: 3
Accuracy Measures
MAPE: 23.909
MAD: 11.184
MSD: 183.098
281
Row Period Forecast Lower Upper
1 71 44.6667 18.1452 71.1881
2 72 44.6667 18.1452 71.1881
3 73 44.6667 18.1452 71.1881
4 74 44.6667 18.1452 71.1881
5 75 44.6667 18.1452 71.1881
Moving Average
Actual
80
Predicted
70 Forecast
Actual
60 Predicted
Forecast
Sales
50
40 Moving Average
Length: 3
30
MAPE: 23.909
20 MAD: 11.184
MSD: 183.098
0 10 20 30 40 50 60 70
Time
:( )&!آ4". ا
ا4Lك & ا
ر%;& w1;& (
' ب3)
MTB > %MA 'Sales' 4;
SUBC> Center;
SUBC> Forecasts 5.
Executing from file: G:\MTBWIN\MACROS\MA.MAC
Macro is running ... please wait
Moving average
Data Sales
282
Length 70.0000
NMissing 0
Moving Average
Length: 4
Accuracy Measures
MAPE: 21.161
MAD: 9.733
MSD: 146.200
Row Period Forecast Lower Upper
1 71 48 24.3010 71.6990
2 72 48 24.3010 71.6990
3 73 48 24.3010 71.6990
4 74 48 24.3010 71.6990
5 75 48 24.3010 71.6990
Moving Average
82 Actual
Predicted
72 Forecast
Actual
62 Predicted
Forecast
Sales
52
Moving Average
42
Length: 4
32 MAPE: 21.161
MAD: 9.733
22 MSD: 146.200
0 10 20 30 40 50 60 70
Time
:4 &[
ا4Lك & ا
ر%;& w1;& ( أا
ب4)
MTB > %MA 'Sales' 5;
SUBC> Forecasts 5.
283
Executing from file: G:\MTBWIN\MACROS\MA.MAC
Macro is running ... please wait
Moving average
Data Sales
Length 70.0000
NMissing 0
Moving Average
Length: 5
Accuracy Measures
MAPE: 21.724
MAD: 10.040
MSD: 156.195
Row Period Forecast Lower Upper
1 71 46.6 22.1043 71.0957
2 72 46.6 22.1043 71.0957
3 73 46.6 22.1043 71.0957
4 74 46.6 22.1043 71.0957
5 75 46.6 22.1043 71.0957
284
Moving Average
Actual
80
Predicted
70 Forecast
Actual
Predicted
60
Forecast
Sales
50
Moving Average
40 Length: 5
30 MAPE: 21.724
MAD: 10.040
20 MSD: 156.195
0 10 20 30 40 50 60 70
Time
ـ4!B HB أ76"2 ا
!!آ4". ا
ا4Lط & ا
ر%;!
اw1;!
أن ا4J.
_ ا8;'
& ا
146.2 ويC (MSD (Mean Square Deviation
7 ه95% :('C ;اتN V& 4(J; !
) اB 4 ![
ات:(';
ا
Row Period Forecast Lower Upper
1 71 48 24.3010 71.6990
2 72 48 24.3010 71.6990
3 73 48 24.3010 71.6990
4 74 48 24.3010 71.6990
5 75 48 24.3010 71.6990
)5
! ا5
) ا ا.
ث ا
"!ت1%. ء و59 ) اB
م1"
ا4آ
د1" f!
ا4"&L
285
ا(;9ر ا
]
! 7ل ا
H bا
]
1422/1421 7هـ
!دة = ) Q%. 221ق ا
;'( :ا( 78 59
ا
& 3 :ت
أ V!L WLا 4ا
;
:4
ا
:ال اول:
'!1ذج
(1 − 1.2 B + 0.6B ) ( z
2
t )− 65) = (1 − 0.4 B ) at , at ~ WN ( 0,1
)أ( & J%Cان ا
'!1ذج & ; Jو J
H.Bب.
)ب( أو Lآ ρ k & Hو . k = 1, 2,...,5 )J
φkk
)ج( أو Lدا
4اوزان . j = 1, 2,...,5 )J
ψ j
ا
:ال ا
]
:7
'!1ذج ا
.إذا ! Sأن . z76 = 60.4, z77 = 58.9, z78 = 64.7, z79 = 70.4, z80 = 62.6
)أ( أو:('C Lات
)Jا
! ;. z81 , z82 , z83 , z84 4(J
)ب( أو;N Lات :(';
95% :('Cات 7Nا
Jbة ا
.4J.
ا
:ال ا
]
:Q
'!1ذج
)(1 − 0.43B )(1 − B ) zt = at , at ~ WN ( 0,1
)أ( ه Hا
'!1ذج & ;J؟ و
!ذا؟
)ب( إذا آ
W 5{N z49 = 33.4, z50 = 33.9 Sا
;'(:ات ) . ℓ = 1, 2,...,5 )J
z50 ( ℓ
)ج( أو;N Lات :(';
95% :('Cات 7Nا
Jbة ا
.4J.
286
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
) Bا 59ء و1%.ث ا
"!ت
آ 4ا
"1م
4"&Lا
!1" fد
ا(;9ر ا
' H b
78Oا
]
1422/1421 7هـ
!دة = ) Q%. 221ق ا
;'( :ا(78 59
ﺍﻝﺯﻤﻥ 3ﺴﺎﻋﺎﺕ
أ V!L WLا 4ا
;
:4
ا
:ال اول:
=("(& 4 ;& 'Jت !1ذج ) ARIMA (1,1,0ا
;
:7
(1 − 0.7 B )(1 − B ) zt = at , )at ∼ WN ( 0, 25
ا
!Kه Cاة هz143 = 770, z144 = 800 7
(أ( ه Hا
'!1ذج & ; Jأم Xو
!ذا؟
)ب( أو Lدا
4اوزان . j = 1, 2,...,5 )J
ψ j
)ج( ا W 5ا
;'(:ات
] ث ) (3ا
)Jا
! ; 4(Jا
;
.4
)د( او;N Lات :(';
95% :('Cات ا
.4J.
ا
:ال ا
]
:7
!; 4ا
&' 4ا
;
) : 4إBأ & ا
ر
! 6ا (6 .
3.49 5.74 5.51 3.99 3.45 4.77 4.14
4.60 3.80 5.43
3.96 2.54 4.05 6.16 3.78 5.07 5.42
3.91 4.30 3.88
2.89 4.61 4.08 4.05 3.28 2.65 1.22
3.98 3.45 3.57
287
2.52 1.58 4.00 5.14 3.84 4.40 3.08
5.43 4.80 2.75
5.77 4.99 4.31 6.46 6.11 4.79 5.65
5.52 6.12 6.06
3.20 5.05 6.23 6.12 4.99 4.89 4.78
5.67 6.08 5.80
5.13 7.07 8.02 6.36 5.75 5.70 5.61
5.63 5.71 5.16
7.20 6.87 7.56 6.57 6.08 4.72 6.09
6.64 7.49 6.64
7.26 7.22 6.69 7.49 9.01 7.27 5.62
7.59 7.53 6.43
6.42 8.22 7.67 7.53 7.23 8.50 8.27
8.75 7.50 7.86
:7
;
اL أوMINITAB ;[امF.
Time Series Plot . ;[امF. 4'&
ا4 ;!
)أ( أر) ا
.)
اo%N ". 4 ;!
اO!;
4J2= ;حB)ب( أ
S
! إذا آN 4Pb
وأ;( اResiduals 7Bا1(
اo%b. )B 45;J& 4J2= H3
()ج
Kolmogorov-Smirnov ;[ام إ;(رF. f
وذ7"(= V2ز1C 4ز1& 7Bا1(
ا
Test .
Mean Absolute Percentage Error ء6
4B
ا2"& رنB ()د
Mean وMean Absolute Deviation (MAD) ( وMAPE)
] اآ4J26
) أ;[م اU & و45;J& 4J2= H3
Squared Deviation (MSD)
95% . :('C ;اتN د2 إV& 4(J; !
) اJ
ا4 ![
ات:('C
1;
4Bد
:Q
]
ال ا:
ا
(6 . ا6 !
أ & ا
رB )إ:4
;
ا4'&
ا4 ;!
1.20 1.50 1.54 2.70 1.95 2.40
3.44 2.83
1.76 2.00 2.09 1.89 1.80 1.25
1.58 2.25
288
2.50 2.05 1.46 1.54 1.42 1.57
1.40 1.51
1.08 1.27 1.18 1.39 1.42 2.08
1.85 1.82
2.07 2.32 1.23 2.91 1.77 1.61
1.25 1.15
1.37 1.79 1.68 1.78 1.84
:7
;
اL أوMINITAB ;[امF.
Time Series Plot . ;[امF. 4'&
ا4 ;!
)أ( أر) ا
ARIMA ( p, d , q ) 48 & W'& ذج1!
;حB أO'& وSPACF وSACF & H آL)ب( أو
. q وd وp & H;" آ. f
وذ
2ة ا
!"
) & وA هH أو آ5! إذا آن أN تPb
;ح وأ;( اJ!
ذج ا1!'
)
"!
ر اB ()ج
.b
V2ز1C 4ز1& 7Bا1(
اS ! إذا آN 4Pb
وأ;( اResiduals 7Bا1(
اo%b. )B ()د
Kolmogorov-Smirnov Test . ;[ام إ;(رF. f
وذ7"(=
95% . :('C ;اتN د2 إV& 4
;
ا4(J; !
) اJ
ا4 ![
ات:('C W 5)هـ( أ
289
)5
! ا5
) ا ا.
هـ1422/1421 7
]
اH b
78O'
;(ر ا4!;%& ت.Lإ
Q%. 221
!دة
:ال اول:
ا4.Lإ
(1 − B ) 2b;
اH& ى1%2 *
J; & z ذج1!'
)أ( ا
()ب
(1 − 0.7 B )(1 − B ) zt = at
(1 − 1.7 B + 0.7 B ) z
2
t = at
1
∴ zt = at
1 − 1.7 B + 0.7 B 2
= ψ ( B ) at
1
∴ψ ( B ) =
1 − 1.7 B + 0.7 B 2
∴ (1 + ψ 1B + ψ 2 B 2 + ψ 3 B 3 + ⋯)(1 − 1.7 B + 0.7 B 2 ) ≡ 0
B : ψ 1 − 1.7 = 0 ⇒ ψ 1 = 1.7
B 2 : ψ 2 − 1.7ψ 1 + 0.7 = 0 ⇒ ψ 2 = 1.7ψ 1 − 0.7 = 2.19
B 3 : ψ 3 − 1.7ψ 2 + 0.7ψ 1 = 0 ⇒ ψ 3 = 1.7ψ 2 − 0.7ψ 1 = 2.53
⋮
B j : ψ j − 1.7ψ j −1 + 0.7ψ j −2 = 0 ⇒ ψ j = 1.7ψ j −1 + 0.7ψ j −2 , j = 2, 3,...
∴ψ 4 = 1.7ψ 3 − 0.7ψ 2 = 2.768
ψ 5 = 1.7ψ 4 − 0.7ψ 3 = 2.9346
ψ 1 = 1.7,ψ 2 = 2.19,ψ 3 = 2.53,ψ 4 = 2.77,ψ 5 = 2.93 :7 ه4.16!
إذا اوزان ا
:7
;
ات آ:(';
اW %
()ج
290
∵ (1 − 1.7 B + 0.7 B 2 ) zt = at
∴ zt = 1.7 zt −1 − 0.7 zt −2 + at
∴ zt ( ℓ ) = E zt + ℓ zt , zt −1 ,⋯ , ℓ ≥ 0
= E 1.7 zt + ℓ−1 − 0.7 zt + ℓ−2 + at + ℓ zt , zt −1 ,⋯ , ℓ ≥ 0
= 1.7 E zt + ℓ−1 zt , zt −1 ,⋯ − 0.7 E zt +ℓ −2 zt , zt −1 ,⋯ + E at +ℓ zt , zt −1 ,⋯ , ℓ ≥ 0
∴ ℓ = 1: zt (1) = 1.7 E zt zt , zt −1 ,⋯ − 0.7 E zt −1 zt , zt −1 ,⋯ + E at +1 zt , zt −1 ,⋯
= 1.7 zt − 0.7 zt −1
ℓ = 2 : zt ( 2 ) = 1.7 E zt +1 zt , zt −1 ,⋯ − 0.7 E zt zt , zt −1 ,⋯ + E at + 2 zt , zt −1 ,⋯
= 1.7 zt (1) − 0.7 zt
ℓ = 3 : zt ( 3) = 1.7 E zt + 2 zt , zt −1 ,⋯ − 0.7 E zt +1 zt , zt −1 ,⋯ + E at +3 zt , zt −1 ,⋯
= 1.7 zt ( 2 ) − 0.7 zt (1)
∴ ℓ ≥ 3 : zt ( ℓ ) = 1.7 zt ( ℓ − 1) − 0.7 zt ( ℓ − 2 )
∵ t = 144
∴ z144 (1) = 1.7 z144 − 0.7 z143 = 1.7 ( 800 ) − 0.7 ( 770 ) = 821
z144 ( 2 ) = 1.7 z144 (1) − 0.7 z144 = 1.7 ( 821) − 0.7 ( 800 ) = 835.7
z144 ( 3) = 1.7 z144 ( 2 ) − 0.7 z144 (1) = 1.7 ( 835.7 ) − 0.7 ( 821) = 845.99
:7 ه4(J; !
) اB ات
] ث:(';
إذا ا
z144 (1) = 821, z144 ( 2 ) = 835.7, z144 ( 3) = 845.99
4B "
. 6"C :('C (1 − α )100% ;اتN ()د
z (ℓ) ± u
et ( ℓ ) , ℓ ≥ 0
α 2 V
t
أي أنuα 2 = 1.96 نFN α = 0.05 انQ5 و7J
ا7"(6
اV2ز1;
95 !
ا1 هuα 2 Q5
zt + ℓ ∈ zt ( ℓ ) ± 1.96 V et ( ℓ ) w.p. 0.95, ℓ ≥ 0
V et ( ℓ ) = σ 2 (1 + ψ 12 + ψ 22 + ⋯ + ψ ℓ2−1 ) , ℓ ≥ 0 :4B "
& ا:(';
ء ا6'ت أ2(C W % Xأو
V et (1) = σ 2 = 25
(
V et ( 2 ) = σ 2 (1 + ψ 12 ) = 25 1 + (1.7 ) = 97.25
2
)
(
V et ( 3) = σ 2 (1 + ψ 12 + ψ 22 ) = 25 1 + (1.7 ) + ( 2.19 ) = 217.1525
2 2
)
:4.16!
ا:(';
;ات اN
O'&و
291
z145 ∈ 821 ± 1.96 25 = [811.2,830.8] , w. p. 0.95
z146 ∈ 835.7 ± 1.96 97.25 = [816.37,855.03] , w. p. 0.95
z145 ∈ 845.99 ± 1.96 217.1525 = [817.11,874.87] , w. p. 0.95
:7
]
ال ا:
4.Lإ
( )أ
5
Sales
1
Index 10 20 30 40 50
ق6
ا5ه أO!C 7N
;[مf
A
و4!1& z 4 ;!
أن ا.
)ب( & ا
) ا
:4
;
ا
Moving Average Smoothing ك%;!
اw1;!
ا-1
Single Exponential Smoothing w (
ا7 اO!;
ا-2
Double Exponential Smoothing 78']
ا7 اO!;
ا-3
ك%;!
اw1;!
اXأو
292
Smoothing Sales Series by Moving Avg. of Order 3
Actual
8
Predicted
7 Forecast
Actual
6 Predicted
Forecast
Sales
5
4 Moving Average
Length: 3
3
2 MAPE: 26.5366
MAD: 0.9077
1 MSD: 1.2322
0 10 20 30 40 50
Time
MTB > %MA 'Sales' 3;
SUBC> Forecasts 5;
SUBC> Title "Smoothing Sales Series by Moving
Avg. of Order 3";
SUBC> Residuals 'RESI1'.
Executing from file: G:\MTBWIN\MACROS\MA.MAC
Macro is running ... please wait
Moving average
Data Sales
Length 50.0000
NMissing 0
Moving Average
Length: 3
Accuracy Measures
MAPE: 26.5366
MAD: 0.9077
293
MSD: 1.2322
Row Period Forecast Lower Upper
1 51 5.9 3.72428 8.07572
2 52 5.9 3.72428 8.07572
3 53 5.9 3.72428 8.07572
4 54 5.9 3.72428 8.07572
5 55 5.9 3.72428 8.07572
w (
ا7 اO!;
اU
Smoothing Sales Series by Single Exponential Smoothing
8 Actual
Predicted
7 Forecast
Actual
6 Predicted
Forecast
Sales
4
Smoothing Constant
Alpha: 0.258
3
2 MAPE: 25.5159
MAD: 0.9002
1 MSD: 1.1638
0 10 20 30 40 50
Time
MTB > %SES 'Sales';
SUBC> Forecasts 5;
SUBC> Title "Smoothing Sales Series by Single
Exponential Smoothing";
SUBC> Residuals 'RESI2'.
Executing from file: G:\MTBWIN\MACROS\SES.MAC
Macro is running ... please wait
294
Single Exponential Smoothing
Data Sales
Length 50.0000
NMissing 0
Smoothing Constant
Alpha: 0.257773
Accuracy Measures
MAPE: 25.5159
MAD: 0.9002
MSD: 1.1638
Row Period Forecast Lower Upper
1 51 5.67586 3.47035 7.88137
2 52 5.67586 3.47035 7.88137
3 53 5.67586 3.47035 7.88137
4 54 5.67586 3.47035 7.88137
5 55 5.67586 3.47035 7.88137
78']
ا7 اO!;
] اU
295
Smoothing Sales Series by Double Exponential Smoothing
Actual
11
Predicted
Forecast
Actual
Predicted
Forecast
Sales
6
Smoothing Constants
Alpha (level): 0.681
Gamma (trend):0.019
MAPE: 27.1322
MAD: 0.9880
1 MSD: 1.5458
0 10 20 30 40 50
Time
MTB > %DES 'Sales';
SUBC> Forecasts 5;
SUBC> Title "Smoothing Sales Series by Double
Exponential Smoothing";
SUBC> Residuals 'RESI3'.
Executing from file: G:\MTBWIN\MACROS\DES.MAC
Macro is running ... please wait
Double Exponential Smoothing
Data Sales
Length 50.0000
NMissing 0
Smoothing Constants
Alpha (level): 0.680728
Gamma (trend): 0.019421
Accuracy Measures
MAPE: 27.1322
MAD: 0.9880
296
MSD: 1.5458
Row Period Forecast Lower Upper
1 51 6.04349 3.62298 8.4640
2 52 6.06254 3.04085 9.0842
3 53 6.08160 2.40519 9.7580
4 54 6.10065 1.74005 10.4613
5 55 6.11971 1.05736 11.1821
:7Bا1(
اo%N ()ج
ك%;!
اw1;!
Autocorrelation Function for RESI1
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag Corr T LBQ Lag Corr T LBQ
1 0.01 0.05 0.00 8 0.08 0.47 9.66
2 -0.33 -2.26 5.55 9 -0.11 -0.66 10.43
3 -0.11 -0.69 6.19 10 0.00 0.02 10.43
4 -0.01 -0.06 6.19 11 0.04 0.25 10.55
5 0.10 0.64 6.79
6 0.20 1.22 9.03
7 -0.07 -0.39 9.28
297
Partial Autocorrelation Function for RESI1
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag PAC T Lag PAC T
1 0.01 0.05 8 0.25 1.74
2 -0.33 -2.26 9 -0.11 -0.76
3 -0.12 -0.81 10 0.15 1.01
4 -0.14 -0.93 11 -0.06 -0.39
5 0.03 0.19
6 0.17 1.14
7 -0.02 -0.16
10
Frequency
-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
RESI1
Normal Probability Plot for RESI1
99
Mean: 3.15E-02
StDev: 1.12161
95
90
80
70
Percent
60
50
40
30
20
10
-3 -2 -1 0 1 2 3
Data
MTB > %Qqplot 'RESI1';
SUBC> Table;
298
SUBC> Conf 95;
SUBC> Ci.
Executing from file: G:\MTBWIN\MACROS\Qqplot.MAC
Distribution Function Analysis
Normal Dist. Parameter Estimates
Data : RESI1
Mean: 3.15E-02
StDev: 1.12161
Percentile Estimates
95% CI 95% CI
Approximate Approximate
P Percentile Lower Limit Upper Limit
0.01 -2.57777 -3.19506 -1.96047
0.02 -2.27202 -2.83741 -1.70663
0.03 -2.07803 -2.61158 -1.54447
0.04 -1.93210 -2.44238 -1.42181
0.05 -1.81339 -2.30524 -1.32155
0.06 -1.71236 -2.18891 -1.23581
0.07 -1.62377 -2.08723 -1.16032
0.08 -1.54445 -1.99647 -1.09244
0.09 -1.47231 -1.91417 -1.03046
0.10 -1.40591 -1.83864 -0.97318
0.20 -0.91248 -1.28563 -0.53934
299
0.30 -0.55668 -0.89868 -0.21469
0.40 -0.25267 -0.57843 0.07309
0.50 0.03149 -0.28917 0.35215
0.60 0.31565 -0.01012 0.64141
0.70 0.61966 0.27767 0.96165
0.80 0.97546 0.60232 1.34860
0.90 1.46889 1.03616 1.90162
0.91 1.53529 1.09344 1.97715
0.92 1.60743 1.15542 2.05945
0.93 1.68675 1.22330 2.15021
0.94 1.77534 1.29879 2.25189
0.95 1.87637 1.38453 2.36822
0.96 1.99508 1.48479 2.50536
0.97 2.14101 1.60745 2.67456
0.98 2.33499 1.76961 2.90038
0.99 2.64074 2.02345 3.25804
Kolmogorov-Smirnov Test for Residuals of MA
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-2 -1 0 1 2
RESI1
Average: 0.0314894 Kolmogorov-Smirnov Normality Test
StDev: 1.12161 D+: 0.069 D-: 0.067 D : 0.069
N: 47 Approximate P-Value > 0.15
w (
ا7 اO!;
300
Autocorrelation Function for RESI2
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 7 12
Lag Corr T LBQ Lag Corr T LBQ
1 -0.02 -0.16 0.03 8 0.17 1.05 9.33
2 -0.21 -1.51 2.48 9 -0.13 -0.79 10.37
3 0.16 1.07 3.87 10 -0.08 -0.47 10.76
4 0.01 0.07 3.87 11 0.13 0.81 11.95
5 0.10 0.68 4.47 12 -0.06 -0.38 12.24
6 0.21 1.40 7.18
7 -0.08 -0.53 7.61
Partial Autocorrelation Function for RESI2
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 7 12
Lag PAC T Lag PAC T
1 -0.02 -0.16 8 0.25 1.76
2 -0.21 -1.51 9 -0.27 -1.89
3 0.15 1.09 10 0.04 0.26
4 -0.03 -0.24 11 -0.11 -0.75
5 0.18 1.28 12 -0.09 -0.62
6 0.20 1.42
7 -0.02 -0.14
9
8
7
6
Frequency
5
4
3
2
1
0
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0
RESI2
301
Normal Probability Plot for RESI2
99
Mean: 9.40E-02
StDev: 1.08561
95
90
80
70
Percent
60
50
40
30
20
10
-3 -2 -1 0 1 2 3
Data
MTB > %Qqplot 'RESI2';
SUBC> Table;
SUBC> Conf 95;
SUBC> Ci.
Executing from file: G:\MTBWIN\MACROS\Qqplot.MAC
Distribution Function Analysis
Normal Dist. Parameter Estimates
Data : RESI2
Mean: 9.40E-02
StDev: 1.08561
Percentile Estimates
95% CI 95% CI
302
Approximate Approximate
P Percentile Lower Limit Upper Limit
0.01 -2.43146 -3.01074 -1.85218
0.02 -2.13552 -2.66609 -1.60495
0.03 -1.94776 -2.44846 -1.44706
0.04 -1.80651 -2.28537 -1.32766
0.05 -1.69162 -2.15318 -1.23006
0.06 -1.59383 -2.04103 -1.14663
0.07 -1.50809 -1.94300 -1.07317
0.08 -1.43131 -1.85549 -1.00713
0.09 -1.36149 -1.77614 -0.94684
0.10 -1.29722 -1.70330 -0.89113
0.20 -0.81962 -1.16979 -0.46946
0.30 -0.47525 -0.79618 -0.15431
0.40 -0.18099 -0.48669 0.12471
0.50 0.09405 -0.20686 0.39496
0.60 0.36908 0.06338 0.67479
0.70 0.66334 0.34241 0.98427
0.80 1.00772 0.65756 1.35789
0.90 1.48531 1.07923 1.89140
0.91 1.54959 1.13494 1.96423
0.92 1.61941 1.19523 2.04359
0.93 1.69618 1.26127 2.13110
0.94 1.78193 1.33473 2.22913
0.95 1.87972 1.41816 2.34128
0.96 1.99461 1.51575 2.47347
0.97 2.13586 1.63516 2.63655
0.98 2.32362 1.79305 2.85419
0.99 2.61955 2.04028 3.19883
303
Kolmogorov-Smirnov Test for Residuals of SES
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-2 -1 0 1 2
RESI2
Average: 0.0940483 Kolmogorov-Smirnov Normality Test
StDev: 1.08561 D+: 0.077 D-: 0.068 D : 0.077
N: 50 Approximate P-Value > 0.15
78']
ا7 اO!;
ا
Autocorrelation Function for RESI3
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 7 12
Lag Corr T LBQ Lag Corr T LBQ
1 -0.22 -1.53 2.50 8 0.16 0.96 15.06
2 -0.29 -1.98 7.12 9 -0.16 -0.89 16.62
3 0.22 1.39 9.81 10 0.01 0.04 16.62
4 -0.07 -0.43 10.10 11 0.19 1.06 18.99
5 0.00 0.03 10.10 12 -0.15 -0.82 20.53
6 0.17 1.03 11.82
7 -0.16 -0.95 13.38
Partial Autocorrelation Function for RESI3
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 7 12
Lag PAC T Lag PAC T
1 -0.22 -1.53 8 0.24 1.69
2 -0.36 -2.52 9 -0.23 -1.61
3 0.07 0.49 10 0.13 0.92
4 -0.11 -0.81 11 0.01 0.04
5 0.07 0.47 12 -0.03 -0.25
6 0.14 1.02
7 -0.05 -0.34
304
9
8
7
6
Frequency
5
4
3
2
1
0
-3 -2 -1 0 1 2 3
RESI3
Normal Probability Plot for RESI3
99
Mean: -3.0E-03
StDev: 1.25593
95
90
80
70
Percent
60
50
40
30
20
10
-3 -2 -1 0 1 2 3
Data
MTB > %Qqplot 'RESI3';
SUBC> Table;
SUBC> Conf 95;
SUBC> Ci.
Executing from file: G:\MTBWIN\MACROS\Qqplot.MAC
Distribution Function Analysis
Normal Dist. Parameter Estimates
305
Data : RESI3
Mean: -3.0E-03
StDev: 1.25593
Percentile Estimates
95% CI 95% CI
Approximate Approximate
P Percentile Lower Limit Upper Limit
0.01 -2.92473 -3.59489 -2.25457
0.02 -2.58237 -3.19618 -1.96855
0.03 -2.36515 -2.94440 -1.78589
0.04 -2.20174 -2.75573 -1.64775
0.05 -2.06882 -2.60279 -1.53485
0.06 -1.95569 -2.47305 -1.43832
0.07 -1.85649 -2.35964 -1.35334
0.08 -1.76767 -2.25840 -1.27694
0.09 -1.68689 -2.16659 -1.20719
0.10 -1.61254 -2.08233 -1.14274
0.20 -1.06001 -1.46512 -0.65491
0.30 -0.66161 -1.03289 -0.29032
0.40 -0.32118 -0.67484 0.03248
0.50 -0.00300 -0.35112 0.34512
0.60 0.31519 -0.03847 0.66885
0.70 0.65562 0.28433 1.02690
0.80 1.05402 0.64892 1.45913
0.90 1.60655 1.13675 2.07634
0.91 1.68090 1.20120 2.16060
0.92 1.76168 1.27095 2.25241
306
0.93 1.85050 1.34735 2.35365
0.94 1.94969 1.43233 2.46706
0.95 2.06283 1.52886 2.59680
0.96 2.19575 1.64176 2.74973
0.97 2.35915 1.77990 2.93840
0.98 2.57637 1.96256 3.19019
0.99 2.91874 2.24858 3.58890
Kolmogorov-Smirnov Test for Residuals of DES
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-3 -2 -1 0 1 2
RESI3
Average: -0.0029958 Kolmogorov-Smirnov Normality Test
StDev: 1.25593 D+: 0.077 D-: 0.070 D : 0.077
N: 50 Approximate P-Value > 0.15
تPb
اJ%C 7Bا1(
ان ا4J.
ت اX%
اH آ7N
46. &;اz -1
ة5 و2(Cي وb+ w1;!. 7"(= V2ز1C (2JC O
-2
ء6
4B
~ ا2J& o[2 7
;
)د( ا
ول ا
MAPE MAD MSD
MA 26.5366 0.9077 1.2322
SES 25.5159 0.9002 1.1638
307
DES 27.1322 0.9880 1.5458
.4 ; H`N أ76"2 w (
ا7 اO!;
أن ا5 2و
:(';
95% ;اتN وw (
ا7 اO!;
;[ام اF. ات:(';
ا
Period of
Forecast Forecast Lower Upper
51 5.67586 3.47035 7.88137
52 5.67586 3.47035 7.88137
53 5.67586 3.47035 7.88137
54 5.67586 3.47035 7.88137
55 5.67586 3.47035 7.88137
:Q
]
ال ا:
4.Lإ
( )أ
3.5
3.0
2.5
Defects
2.0
1.5
1.0
Index 10 20 30 40
308
Autocorrelation Function for Defects
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag Corr T LBQ Lag Corr T LBQ
1 0.43 2.88 8.84 8 -0.11 -0.57 17.25
2 0.26 1.49 12.18 9 -0.05 -0.27 17.41
3 0.14 0.77 13.18 10 -0.01 -0.04 17.41
4 0.08 0.43 13.50 11 -0.04 -0.19 17.50
5 -0.09 -0.46 13.89
6 -0.07 -0.39 14.18
7 -0.21 -1.10 16.57
Partial Autocorrelation Function for Defects
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag PAC T Lag PAC T
1 0.43 2.88 8 0.07 0.44
2 0.09 0.63 9 0.05 0.35
3 -0.00 -0.01 10 0.01 0.09
4 0.00 0.00 11 -0.03 -0.23
5 -0.16 -1.07
6 0.00 0.02
7 -0.18 -1.19
ذج1! ;حJ2 اA وهr1,1 ". V6B O
SPACF ;[& أ وC SACF أنH3K
^ & اPوا
q = 0 وd = 0 وp = 1 أي أنARIMA (1, 0, 0 )
:)
"!
ا2JC ()ج
MTB > ARIMA 1 0 0 'Defects' 'RESI1' 'FITS1';
SUBC> Constant;
SUBC> Forecast 5 c3 c4 c5;
SUBC> GACF;
SUBC> GPACF;
SUBC> GHistogram;
SUBC> GNormalplot;
SUBC> GFits;
SUBC> GOrder.
309
ARIMA Model
ARIMA model for Defects
Estimates at each iteration
Iteration SSE Parameters
0 11.2419 0.100 1.700
1 10.0858 0.250 1.393
2 9.5649 0.400 1.086
3 9.5316 0.436 1.006
4 9.5309 0.441 0.995
5 9.5309 0.442 0.993
6 9.5309 0.442 0.993
Relative change in each estimate less than
0.0010
Final Estimates of Parameters
Type Coef StDev T
AR 1 0.4421 0.1365 3.24
Constant 0.99280 0.06999 14.19
Mean 1.7795 0.1254
Number of observations: 45
Residuals: SS = 9.47811 (backforecasts
excluded)
MS = 0.22042 DF = 43
Modified Box-Pierce (Ljung-Box) Chi-Square
statistic
Lag 12 24
36 48
310
Chi-Square 4.9(DF=11) 8.9(DF=23)
30.9(DF=35) * (DF= *)
Forecasts from period 45
95 Percent Limits
Period Forecast Lower Upper
Actual
46 1.80627 0.88588 2.72665
47 1.79135 0.78503 2.79767
48 1.78476 0.76248 2.80703
49 1.78184 0.75648 2.80721
50 1.78055 0.75459 2.80652
1;ح هJ!
ذج ا1!'
أن ا4J.
ت اL[!
& ا
zt = 0.9928 − 0.4421zt −1 + at , at ∼ WN ( 0, 0.22042 )
or ( zt − 1.7795) = −0.4421( zt −1 − 1.7795) + at
Q5
( )
φˆ1 = 0.4421, s.e φˆ1 = 0.1365, with t-value = 3.24
ي1'"& 7CاA
ار ا%
9 اH&"& أي أن
( )
δˆ = 0.9928, s.e δˆ = 0.06999, with t-value = 14.19
ي1'"& δ ى1; !
` ا2وأ
:7Bا1(
اo%N ()د
w. م ا
;ا-1
311
ACF of Residuals for Defects
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag
PACF of Residuals for Defects
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 2 3 4 5 6 7 8 9 10 11
Lag
:7"(6
اV2ز1;
ا
Histogram of the Residuals
(response is Defects)
10
Frequency
-1.0 -0.5 0.0 0.5 1.0 1.5
Residual
312
Normal Probability Plot of the Residuals
(response is Defects)
Residual
0
-1
-2 -1 0 1 2
Normal Score
Normal Probability Plot for RESI1
99
Mean: 8.21E-03
StDev: 0.464050
95
90
80
70
Percent
60
50
40
30
20
10
-1.0 -0.5 0.0 0.5 1.0 1.5
Data
MTB > %Qqplot 'RESI1';
SUBC> Table;
SUBC> Conf 95;
SUBC> Ci.
Executing from file: G:\MTBWIN\MACROS\Qqplot.MAC
Distribution Function Analysis
Normal Dist. Parameter Estimates
Data : RESI1
313
Mean: 8.21E-03
StDev: 0.464050
Percentile Estimates
95% CI 95% CI
Approximate Approximate
P Percentile Lower Limit Upper Limit
0.01 -1.07134 -1.33235 -0.81033
0.02 -0.94484 -1.18390 -0.70577
0.03 -0.86458 -1.09018 -0.63897
0.04 -0.80420 -1.01996 -0.58844
0.05 -0.75509 -0.96306 -0.54712
0.06 -0.71329 -0.91478 -0.51179
0.07 -0.67663 -0.87260 -0.48067
0.08 -0.64382 -0.83494 -0.45269
0.09 -0.61397 -0.80080 -0.42714
0.10 -0.58650 -0.76947 -0.40353
0.20 -0.38235 -0.54012 -0.22457
0.30 -0.23514 -0.37975 -0.09054
0.40 -0.10936 -0.24710 0.02838
0.50 0.00821 -0.12738 0.14379
0.60 0.12577 -0.01197 0.26351
0.70 0.25155 0.10695 0.39616
0.80 0.39876 0.24098 0.55654
0.90 0.60291 0.41994 0.78588
0.91 0.63038 0.44355 0.81721
0.92 0.66023 0.46911 0.85136
0.93 0.69305 0.49708 0.88901
314
0.94 0.72970 0.52820 0.93120
0.95 0.77150 0.56353 0.97947
0.96 0.82061 0.60485 1.03638
0.97 0.88099 0.65539 1.10659
0.98 0.96125 0.72219 1.20031
0.99 1.08775 0.82674 1.34876
Normal Probability Plot
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-0.5 0.0 0.5 1.0
RESI1
Average: 0.0082067 Kolmogorov-Smirnov Normality Test
StDev: 0.464050 D+: 0.166 D-: 0.080 D : 0.166
N: 45 Approximate P-Value < 0.01
4J(6!
) اJ
اV& 7Bا1(
أ
!ط اo%b
Residuals Versus the Fitted Values
(response is Defects)
1
Residual
-1
1.5 2.0 2.5
Fitted Value
ا
(
تWCC V& 7Bا1(
وا
315
Residuals Versus the Order of the Data
(response is Defects)
Residual
0
-1
5 10 15 20 25 30 35 40 45
Observation Order
ة6"!
ا
(
ت ا7 (6;
W'& ;حJ!
ذج ا1!'
;';_ أن ا4J.
ت ا+1%b
& ا
()هـ
95% Forecasts from period 45 :('C ;اتN و4(J; !
) اJ
ات
[!~ ا:(';
ا
95 Percent Limits
Period Forecast Lower Upper
46 1.80627 0.88588 2.72665
47 1.79135 0.78503 2.79767
48 1.78476 0.76248 2.80703
49 1.78184 0.75648 2.80721
50 1.78055 0.75459 2.80652
316
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
ﻗﺴﻡ ﺍﻹﺤﺼﺎﺀ ﻭﺒﺤﻭﺙ ﺍﻝﻌﻤﻠﻴﺎﺕ
ا
!دة = :ق ا
;'( :اQ%. 221 78 5X
ﺍﻻﺨﺘﺒﺎﺭ ﺍﻻﻭل ﻝﻸﻋﻤﺎل ﺍﻝﻔﺼﻠﻴﺔ
ا
H bاول 1421-1420هـ
ا
& ; :
أ V!L 7 WLا 4ا
;
:4
ا
:ال اول:
ا
H3K 7 z n + j = β 0 + β1 j + ε n + j , j = 0,±1,... ا
;
7 ا
'!1ذج VP ا-
ان Lو.ه βو ) f ′( jأوL z n + j = f ′( j )β + ε n + j , j = 0,±1,...وذ
2%;. fآ& H
) f ′( j + 1) = Lf ′( j
ان ]). X ′ = [f (− n + 1), f (− n + 2 ),..., f (1), f (0ه ب y ′ = (z1 ,..., z n ) VP1. -و
n −1 n −1
) X ′X = ∑ f (− j )f ′(− jوان X ′y = ∑ f (− j )z n − jو& )UاوJ& Lر
!") β
j =0 j =0
ا
:ال ا
]
:7
!; 4ا
&' 4ا
;
:4
t 1 2 3 4 5 6 7 8 9 10 11 12
zt 7 8 12 13 17 15 19 22 25 24 26 28
ا (= -و%;& wك & ا
ر 4 ;!
3 4Lا
!" @6و& )Uاو(';& Lت
)Jا
!; 4
z13 , z14
ب O!C (= α = 0.5 A{. -ا 4 ;!
w . 7ا
!" @6و& )Uاو(';& Lت
)Jا
!; 4
z13 , z14
317
:Q
]
ال ا:
ا
و φ1 < 1 وS.U ارJ& 0 < δ < ∞ Q5 z t = δ + φ1 z t −1 + at 7
;
ذج ا1!'
at ~ WN (0, σ 2 )
k = 0,1,...,5 )J
O! وارρ k , ∀k 7CاA
اw. ا
;ا4
داL او-ا
k = 0,1,...,5 )J
O! و وارφk ,k , ∀k 78
ا7CاA
اw. ا
;ا4
داL او-ب
318
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
4"&Lا
!1" fد – آ 4ا
"1م
) Bا 59ء و1%.ث ا
"!ت
إ;(ر أ!ل H Nأول 1423/1422هـ
!دة =) Q%. 221ق ا
;'( :ا78 59
ﺍﻝﺯﻤﻥ 2 :ﺴﺎﻋﺔ
أ V!L WLا 4ا
;
:4
ا
:ال اول:
'!1ذج
z = β0 + β1t + x t
) x t = φ1x t −1 + at , at ∼ WN (0, σ 2
إ;Nاض أن ا
!"
) .4&1"& β0 , β1, φ1, σ 2
.ه أن ا
!;'(| ا
[ 76ذا أد
&;16[
{6 V.& w1ة ℓإ
ا&م 4B "
. 6"2
zt (ℓ ) = β0 + β1 (t + ℓ ) + φ1ℓ (zt − β0 − β1t ), ℓ≥0
ا
:ال ا
]
:7
إ;!دا &; K& n = 200 O
1= 4هة =(
!1ذج ) AR (2و ' 5ا
;ا6.ت
ا
Aا1(
4Cا 7Bا
;
r1 = 0.13, r2 = 0.13, r3 = 0.12 4إذا آ
φˆ1 = 1.1, φˆ2 = −0.8 S
HONا
;ا6.ت ا
Aا1(
4Cا )C 7Bأن ا
'!1ذج ه AR (2) 1آ5 Hة أو & ;!"4؟
ا
:ال ا
]
:Q
!; 4ا
! ;Jة ا
;
B 6, 5, 4, 6, 4 4ر µ, γ 0, ρ1
319
:V.ال ا
ا:
ا
4
;
!ت ا3
اS( 5 هةK& n = 100 O
1= 4'& ز4 ;!
r1 = 0.8, r2 = 0.5, r3 = 0.4, z = 2, s 2 = 5
H3K
اAR (2) ذج1!
& هاتK!
;ض ان اNإذا ا
z = δ + φ1z t −1 + φ2z t −2 + at , at ∼ WN (0, σ 2 )
φ1, φ2, δ, σ 2 )
"!
رات ا
"ومJ& Lأو
320
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
4"&Lا
!1" fد – آ 4ا
"1م
) Bا 59ء و1%.ث ا
"!ت
ا(;9ر ا
' H b
78Oاول 1423/1422هـ
!دة =) Q%. 221ق ا
;'( :ا(78 59
ﺍﻝﺯﻤﻥ 3 :ﺴﺎﻋﺎﺕ
أ V!L WLا 4ا
;
:4
ا
:ال اول:
'!1ذج
z = β0 + β1t + x t
) x t = φ1x t −1 + at , at ∼ WN (0, σ 2
إ;Nاض أن ا
!"
) .4&1"& β0 , β1, φ1, σ 2
.ه أن ا
!;'(| ا
[ 76ذا أد
&;16[
{6 V.& w1ة ℓإ
ا&م 4B "
. 6"2
z t (ℓ ) = β0 + β1 (t + ℓ ) + φ1ℓ (z t − β0 − β1t ), ℓ≥0
ا
:ال ا
]
:7
إ;!دا &; K& n = 200 O
1= 4هة =(
!1ذج ) AR (2و ' 5ا
;ا6.ت
ا
Aا1(
4Cا 7Bا
;
r1 = 0.13, r2 = 0.13, r3 = 0.12 4إذا آ
φˆ1 = 1.1, φˆ2 = −0.8 S
HONا
;ا6.ت ا
Aا1(
4Cا )C 7Bأن ا
'!1ذج ه AR (2) 1آ5 Hة أو & ;!"4؟
ا
:ال ا
]
:Q
!; 4ا
! ;Jة ا
;
B 6, 5, 4, 6, 4 4ر µ, γ 0, ρ1
ا
:ال ا
ا:V.
!; 4ز&'K& n = 100 O
1= 4هة S( 5ا
!3ت ا
;
4
321
r1 = 0.8, r2 = 0.5, r3 = 0.4, z = 2, s 2 = 5
إذا ا;Nض ان ا
!Kهات &
!1ذج ) AR (2ا
H3K
) z = δ + φ1z t −1 + φ2z t −2 + at , at ∼ WN (0, σ 2
أوJ& Lرات ا
"وم
!"
) φ1, φ2, δ, σ 2
ا
:ال ا
[&~:
و Lأن ا
!("ت ا
' 2 !. 421ا
X2ت
Kآ V(;C & 4ا
'!1ذج
) zt = 5 + 1.1zt −1 − 0.5zt −2 + at , at ∼ WN ( 0, 2
إذا آ
Sا
!("ت
'1ات 1419و 1420و 1421هـ ه 7ا
;1ا
10 7و 11و 2 & 9
ر2ل
-1أو:('C Lات
!("ت 1422و 1423و 1424هـ
ψj , j = 1,2, 3, 4. -2أ W 5اوزان
ج( أ;N W 5ات "(!
95% :('Cت
'1ات 1422و 1423و 1424هـ.
322
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
4"&Lا
!1" fد – آ 4ا
"1م
) Bا 59ء و1%.ث ا
"!ت
ا(;9ر ا
' H b
78Oاول 1423/1422هـ
!دة =) Q%. 221ق ا
;'( :ا(78 59
ﺍﻝﺯﻤﻥ 3 :ﺴﺎﻋﺎﺕ
أ V!L WLا 4ا
;
:4
ا
:ال اول:
'!1ذج
z t = β0 + β1t + x t
) x t = φ1x t −1 + at , at ∼ WN (0, σ 2
إ;Nاض أن ا
!"
) .4&1"& β0 , β1, φ1, σ 2
.ه أن ا
!;'(| ا
[ 76ذا أد
&;16[
{6 V.& w1ة ℓإ
ا&م 4B "
. 6"2
z t (ℓ ) = β0 + β1 (t + ℓ ) + φ1ℓ (z t − β0 − β1t ), ℓ≥0
ا
:ال ا
]
:7
إ;!دا &; K& n = 200 O
1= 4هة =(
!1ذج ) AR (2و ' 5ا
;ا6.ت
ا
Aا1(
4Cا 7Bا
;
r1 = 0.13, r2 = 0.13, r3 = 0.12 4إذا آ
φˆ1 = 1.1, φˆ2 = −0.8 S
HONا
;ا6.ت ا
Aا1(
4Cا )C 7Bأن ا
'!1ذج ه AR (2) 1آ5 Hة أو & ;!"4؟
ا
:ال ا
]
:Q
!; 4ز&'K& n = 100 O
1= 4هة S( 5ا
!3ت ا
;
4
r1 = 0.8, r2 = 0.5, r3 = 0.4, z = 2, s 2 = 5
إذا ا;Nض ان ا
!Kهات &
!1ذج ) AR (2ا
H3K
323
) z = δ + φ1z t −1 + φ2z t −2 + at , at ∼ WN (0, σ 2
أوJ& Lرات ا
"وم
!"
) φ1, φ2 , δ, σ 2
ا
:ال ا
ا:V.
و Lأن ا
!("ت ا
' 2 !. 421ا
X2ت
Kآ V(;C & 4ا
'!1ذج
) zt = 5 + 1.1zt −1 − 0.5zt −2 + at , at ∼ WN ( 0, 2
إذا آ
Sا
!("ت
'1ات 1419و 1420و 1421هـ ه 7ا
;1ا
10 7و 11و 2 & 9
ر2ل
-3أو:('C Lات
!("ت 1422و 1423و 1424هـ
ψj , j = 1,2, 3, 4. -4أ W 5اوزان
ج( أ;N W 5ات "(!
95% :('Cت
'1ات 1422و 1423و 1424هـ.
324
)5
! ا5
) ا ا.
هـ1423/1422 اولH b
78O'
;(ر ا4Lذ1!
ت.Lإ
( 78 59 ا:(';
) =ق اQ%. 221
!دة
:ال اول:
4.Lإ
H3K
ذج ا1!'
4N"!
ت اX ا
!"دV`
z t = β0 + β1t + φ1x t −1 + at , at ~ WN (0, σ 2 )
4B "
& ا6"2 إ
ا&مℓ ة16[
{6 V.& w1;&
ذا أد76[
ا
!;'(| ا
zt (ℓ ) = E (z t +ℓ | zt , z t −1,...), ℓ ≥ 0
∵ x t = z t − β0 − β1t
∴ z t − β0 − β1t = φ1 [z t −1 − β0 − β1 (t − 1)] + at , at ~ WN (0, σ 2 )
zt (ℓ ) − β0 − β1 (t + ℓ ) = E (φ1 [zt +ℓ−1 − β0 − β1 (t + ℓ − 1)] + at +ℓ ) | z t , zt −1,... , ℓ ≥ 0
= E (φ1 [zt +ℓ−1 − β0 − β1 (t + ℓ − 1)] | zt , zt −1,...) + E (at +ℓ | z t , z t −1,...), ℓ ≥ 0
ℓ = 1 : z t (1) − β0 − β1 (t + 1) = E (φ1 (z t − β0 − β1t ) | zt , zt −1,...) + 0
= φ1 (z t − β0 − β1t )
ℓ = 2 : zt (2) − β0 − β1 (t + 2) = φ1 [zt (1) − β0 − β1 (t + 1)] = φ12 (z t − β0 − β1t )
ℓ = 3 : zt (3) − β0 − β1 (t + 3) = φ1 [z t (2) − β0 − β1 (t + 2)] = φ13 (z t − β0 − β1t )
ℓ = 4 : z t (4) − β0 − β1 (t + 4) = φ1 [zt (3) − β0 − β1 (t + 3)] = φ14 (zt − β0 − β1t )
مH3K. اA3وه
z t (ℓ ) − β0 − β1 (t + ℓ ) = φ1 [z t ( ℓ − 1) − β0 − β1 (t + ℓ − 1)] = φ1ℓ (z t − β0 − β1t )
ود%
اWCC ". أو
z t (ℓ ) = β0 + β1 (t + ℓ ) + φ1ℓ (z t − β0 − β1t ), ℓ ≥ 0
.ب16!
ا1وه
325
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
ﺟﺎﻣﻌﺔ ﺍﻟﻤﻠﻚ ﺳﻌﻮﺩ
ﻗﺴﻢ ﺍﻹﺣﺼﺎﺀ ﻭﺑﺤﻮﺙ ﺍﻟﻌﻤﻠﻴﺎﺕ
&دة =) Q%. 221ق ا
;'( :ا( 78 59
ا(;9ر ا
' H b
78Oا
] 1423/1422 7هـ
ا
& 3ت
أ V!L WLا 4ا
;
:4
ا
:ال اول:
'!1ذج
)zt = 200 + 1.2 zt −1 − 0.7 zt − 2 + at + 0.5at −1 , at ∼ N ( 0,1
-1أو Lدا
4ا
;ا w.ا
Aا k = 1, 2,...,5 )J
ρ k 7Cوار!.O
-2أو Lدا
4ا
;ا w.ا
Aا 7Cا
k = 1, 2,...,5 )J
φkk 78وار!.O
-3أو Lدا
4اوزان j = 1, 2,...,5 )J
ψ jوار!.O
-4أو Lدا
4ا
;'(. zn ( ℓ ) , ℓ ≥ 0 :
ا
:ال ا
]
:7
ا
!Kهات ا
;
4 ;!
4ز&' ) :4إBأ & ا
ر 6ا ( 6 .
201 202 203 202 200 198 197 197
198 200 201 199
326
196 193 195 197 199 201 201 201
203 203 200 197
194 195 197 201 204 204 202 200
201 200 198 198
198 198 196 193 194 199 204 206
206 203 200 200
202 204 206 205 202 198 199 201
201 201 204 205
205 201 198 197
:7
;
اWL وأMINITAB _&
. 7N هاتK!
اHاد
;[امF. f
هات وذK!
( ا62 ARIMA 48 & W'& ذج1!
"فC -1
AIC ( m ) = n ln σ a2 + 2m :4B "
. 6"2 يA
واAIC 7CاA
&ت ا1"!
&"ر ا
.4('!
;(رات ا9 اV!L O 2 & 7Bا1(
اo%b. )B ;حJ!
ذج ا1!'
-2
. 95% :('C ;اتN V& مJC 4'& أز8 ;5 4(J; !
) اJ
ات:('C
;ح وJ!
ذج ا1!'
& ا-3
327
ﺑﺴﻢ ﺍﷲ ﺍﻟﺮﺣﻤﻦ ﺍﻟﺮﺣﻴﻢ
هـ1423/1422 7
]
اH b
78O'
;(ر ا4!;%& ت.Lإ
( 78 59 ا:(';
)=ق اQ%. 221 &دة
:ال اول:
4.Lإ
تNا%
إH3 zt = 200 + 1.2 zt −1 − 0.7 zt −2 + at + 0.5at −1 , at ∼ N ( 0,1) ذج1!'
اV`
δ 200
ايµ = = = 400 w1;!
ا
(1 − φ1 − φ2 ) 1 + 0.7 − 1.2
zt − 400 = 1.2 ( zt −1 − 400 ) − 0.7 ( zt − 2 − 400 ) + at + 0.5at −1 , at ∼ N ( 0,1)
:7
;
آρk L1
-1
E {( zt − 400 )( zt − k − 400 )} − 1.2 E {( zt −1 − 400 )( zt − k − 400 )} + 0.7 E {( zt − 2 − 400 )( zt − k − 400 )} =
E {at ( zt − k − 400 )} + 0.5 E {at −1 ( zt − k − 400 )}
γ k − 1.2γ k −1 + 0.7γ k −2 = E {at ( zt −k − 400 )} + 0.5E {at −1 ( zt −k − 400 )}
k = 0 : γ 0 − 1.2γ 1 + 0.7γ 2 = E {at ( zt − 400 )} + 0.5 E {at −1 ( zt − 400 )}
= σ 2 + 0.5 ( 0.7σ 2 ) = 1.35σ 2
k = 1: γ 1 − 1.2γ 0 + 0.7γ 1 = E {at ( zt −1 − 400 )} + 0.5 E {at −1 ( zt −1 − 400 )}
= 0.5σ 2
k = 2 : γ 2 − 1.2γ 1 + 0.7γ 0 = E {at ( zt − 2 − 400 )} + 0.5 E {at −1 ( zt − 2 − 400 )}
=0
k ≥ 2 : γ k − 1.2γ k −1 + 0.7γ k − 2 = 0
) σ 2 = 1 VP1.)
ان4J.
ت اB "
& ا
ρ1 = 0.74436
γ 0 4! J
. اة و4B "
و& ا
γ k = 1.2γ k −1 − 0.7γ k −2 , k = 2,3,...
ρ k = 1.2 ρ k −1 − 0.7 ρ k −2 , k = 2,3,...
∴ ρ 2 = 1.2 ρ1 − 0.7 ρ0 = 1.2 ( 0.74436 ) − 0.7 (1) = 0.193232
328
ρ3 = 1.2 ρ 2 − 0.7 ρ1 = -0.289173
ρ 4 = -0.482271
ρ5 = -0.376304
ρ6 = -0.113975
ρ7 = 0.126642
ρ8 = 0.231754
ρ9 = 0.189455
ρ10 = 0.651180
:7
;
) آCو
ACF of the Model
0.5
ACF
0.0
-0.5
0 1 2 3 4 5 6 7 8 9 10
Lag
φkk L1 -2
329
φ00 = 1, by definition
φ11 = ρ1 = 0.744361, by definition
k −1
ρ k − ∑ φk −1, j ρ k − j
j =1
φkk = k −1
, k = 2,3,...
1 − ∑ φk −1, j ρ j
j =1
φkj = φk −1, j − φkkφk −1,k −1 , j = 1, 2,..., k − 1
ρ 2 − φ11 ρ1 0.193233 − ( 0.744361)( 0.744361) −0.3608402
φ22 = = = = −0.8091915
1 − φ11 ρ1 1 − ( 0.744361)( 0.744361) 0.4459268
φ33 = 0.343852
φ44 = -0.165706
φ55 = 0.821095
φ66 = -0.409620
φ77 = 0.204689
φ88 = -0.102327
φ99 = 0.511617
φ10,10 = -0.255822
:7
;
آO!
و
PACF of the Model
0.8
0.6
0.4
0.2
PACF
0.0
-0.2
-0.4
-0.6
-0.8
0 1 2 3 4 5 6 7 8 9 10
Lag
ψ j اوزان4
دا-3
330
:H3K
ذج ا1!'
اV`
zt = 200 + 1.2 zt −1 − 0.7 zt − 2 + at + 0.5at −1
zt − 1.2 zt −1 + 0.7 zt − 2 = 200 + at + 0.5at −1
(1 − 1.2B + 0.7 B ) z
2
t = 200 + (1 + 0.5 B ) at
zt =
200
+
(1 + 0.5B ) a
1 − 1.2 + 0.7 (1 − 1.2 B + 0.7 B 2 )
t
= 400 + ψ ( B ) at
:7 اوزان ه4
دا
ψ (B) =
(1 − 0.5B ) = 1 + ψ 1B + ψ 2 B 2 + ψ 3 B 3 + ...
(1 − 1.2B + 0.7 B 2 )
:7
;
اوزان آL1
(1 + ψ B + ψ
1 2 B 2 + ψ 3 B 3 + ...)(1 − 1.2 B + 0.7 B 2 ) ≡ (1 + 0.5B )
B :ψ 1 − 1.2 = 0.5 ⇒ ψ 1 = 1.7
B 2 :ψ 2 − 1.2ψ 1 + 0.7 = 0 ⇒ ψ 2 = 1.2ψ 1 − 0.7 = 1.34
B 3 :ψ 3 − 1.2ψ 2 + 0.7ψ 1 = 0 ⇒ ψ 3 = 1.2ψ 2 − 0.7ψ 1 = 0.418
⋮
B j :ψ j = 1.2ψ j −1 − 0.7ψ j − 2
ψ 4 = 1.2ψ 3 − 0.7ψ 2 = -0.4364
ψ 5 = -0.81628
ψ 6 = -0.674056
ψ 7 = -0.237471
ψ 8 = 0.186874
ψ 9 = 0.390478
ψ 10 = 0.337762
:7
;
اH3K
اO
و
331
Psi Weights of the Model
1
Psi
-1
0 5 10
:4B "
. 6"C zn ( ℓ ) , ℓ ≥ 0 :(';
ا4
دا-4
zn ( ℓ ) = E ( zn + ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
V&
a , j ≤ 0
E ( an + j z n , z n −1 ,⋯) = n + j
0, j>0
zn+ j , j≤0
E ( zn + j zn , z n −1 ,⋯) =
zn ( j ) , j > 0
:إذا
zn ( ℓ ) = E ( zn + ℓ zn , zn −1 ,⋯) , ℓ ≥ 1
= E ( 200 + 1.2 zn + ℓ −1 − 0.7 zn + ℓ − 2 + an + ℓ + 0.5an + ℓ −1 | zn , zn −1 ,⋯) , ℓ ≥ 1
= 200 + 1.2 E ( zn + ℓ −1 | zn , zn −1 ,⋯) − 0.7 E ( zn + ℓ − 2 | zn , zn −1 ,⋯)
+ E ( an + ℓ | zn , zn −1 ,⋯) + 0.5 E ( an + ℓ −1 | zn , zn −1 ,⋯) , ℓ ≥ 1
ℓ = 1: zn (1) = 200 + 1.2 zn − 0.7 zn −1 + 0.5an
ℓ = 2 : zn ( 2 ) = 200 + 1.2 zn (1) − 0.7 zn
ℓ ≥ 3 : zn ( ℓ ) = 200 + 1.2 zn ( ℓ − 1) − 0.7 zn ( ℓ − 2 )
:4
داH3K.و
332
200 + 1.2 zn − 0.7 zn −1 + 0.5an , ℓ =1
zn ( ℓ ) = 200 + 1.2 zn (1) − 0.7 zn , ℓ=2
200 + 1.2 z ( ℓ − 1) − 0.7 z ( ℓ − 2 ) , ℓ ≥ 3
n n
:7
]
ال ا:
4.Lإ
4 ;!
) ا
205
Observed Series
200
195
Index 10 20 30 40 50 60
2(;
واVB1;
ا7N ةJ; & O
(و ا2
78
ا7CاA
اw. وا
;ا7CاA
اw. ا
;ا7;
) آ & دا
333
Autocorrelation Function for Observed
1.0
Autocorrelation
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 6 11 16
Lag Corr T LBQ Lag Corr T LBQ Lag Corr T LBQ
1 0.77 6.18 39.96 8 0.11 0.55 56.38 15 0.16 0.75 60.52
2 0.31 1.67 46.47 9 0.08 0.38 56.84 16 0.13 0.63 62.00
3 -0.09 -0.48 47.06 10 -0.00 -0.01 56.84
4 -0.26 -1.33 51.72 11 -0.07 -0.33 57.21
5 -0.21 -1.06 54.89 12 -0.06 -0.27 57.47
6 -0.06 -0.32 55.19 13 0.02 0.08 57.49
7 0.06 0.29 55.45 14 0.11 0.52 58.44
Partial Autocorrelation Function for Observed
Partial Autocorrelation
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
1 6 11 16
Lag PAC T Lag PAC T Lag PAC T
1 0.77 6.18 8 0.05 0.44 15 0.01 0.04
2 -0.71 -5.68 9 -0.11 -0.87 16 -0.07 -0.56
3 0.24 1.91 10 0.01 0.06
4 0.07 0.56 11 0.05 0.42
5 -0.08 -0.67 12 0.10 0.77
6 0.08 0.62 13 -0.04 -0.35
7 -0.04 -0.34 14 0.09 0.74
& 41! & (6 ARMA ( p, q ) H3K
ا7ة وهJ; & 4 ;!
( ان اC ;
آ & ا
ا
)J
ا
'!ذج
:7
;
آAIC ( m ) = n ln σ a2 + 2m
!"ر4!B HB اA{ وp = 0,1, 2 q = 0,1, 2
334
ARMA (1, 0 ) 1-
Type Coef StDev T
AR 1 0.7841 0.0803 9.77
Constant 43.1776 0.2587 166.90
Mean 199.954 1.198
Number of observations: 64
Residuals: SS = 264.844 (backforecasts
excluded)
MS = 4.272 DF = 62
= 64 ln(4.272)+2(3) = 64(1.452)+6 = 98.928 AIC ( m ) = n ln σ a2 + 2m
ARMA ( 2, 0 ) 2-
Type Coef StDev T
AR 1 1.3568 0.0870 15.60
AR 2 -0.7422 0.0872 -8.51
Constant 77.0740 0.1768 435.96
Mean 200.013 0.459
Number of observations: 64
Residuals: SS = 121.868 (backforecasts
excluded)
MS = 1.998 DF = 61
= 64 ln(1.998)+2(4) = 64(0.692)+8 = 52.288 AIC ( m ) = n ln σ a2 + 2m
ARMA ( 0,1) 3-
Type Coef StDev T
MA 1 -0.8772 0.0775 -11.32
Constant 200.032 0.464 430.96
Mean 200.032 0.464
Number of observations: 64
335
Residuals: SS = 252.640 (backforecasts
excluded)
MS = 4.075 DF = 62
= 64 ln(4.075) + 2(3) = 64(1.40487) + 6 = 95.912 AIC ( m ) = n ln σ a2 + 2m
ARMA ( 0, 2 ) 4-
Type Coef StDev T
MA 1 -1.3321 0.1042 -12.78
MA 2 -0.6491 0.1032 -6.29
Constant 200.038 0.580 344.69
Mean 200.038 0.580
Number of observations: 64
Residuals: SS = 149.291 (backforecasts
excluded)
MS = 2.447 DF = 61
= 64 ln(2.447) + 2(4) = 64(0.89486) + 8 = 65.271 AIC ( m ) = n ln σ a2 + 2m
ARMA (1,1) 5-
Type Coef StDev T
AR 1 0.6823 0.1023 6.67
MA 1 -0.6832 0.1054 -6.48
Constant 63.5287 0.3342 190.09
Mean 199.954 1.052
Number of observations: 64
Residuals: SS = 153.610 (backforecasts
excluded)
MS = 2.518 DF = 61
= 64 ln(2.518) + 2(4) = 64(0.92346) + 8 = 67.101 AIC ( m ) = n ln σ a2 + 2m
ARMA ( 2,1) 6-
Model cannot be estimated with these data
336
ARMA (1, 2 ) 7-
Type Coef StDev T
AR 1 0.5112 0.1492 3.43
MA 1 -1.0134 0.1518 -6.68
MA 2 -0.4821 0.1482 -3.25
Constant 97.7668 0.4536 215.55
Mean 200.014 0.928
Number of observations: 64
Residuals: SS = 126.688 (backforecasts
excluded)
MS = 2.111 DF = 60
= 64 ln(2.111) + 2(5) = 64(0.74716) + 10 = AIC ( m ) = n ln σ a2 + 2m
57.818
ARMA ( 2, 2 ) 8-
Type Coef StDev T
AR 1 1.1047 0.2011 5.49
AR 2 -0.5789 0.1608 -3.60
MA 1 -0.4394 0.2275 -1.93
MA 2 -0.1918 0.1973 -0.97
Constant 94.8688 0.2824 335.90
Mean 200.041 0.596
Number of observations: 64
Residuals: SS = 113.051 (backforecasts
excluded)
MS = 1.916 DF = 59
= 64 ln(1.916) + 2(6) = 64(0.65) + 12 = 53.6 AIC ( m ) = n ln σ a2 + 2m
:7
;
ا
ول ا7N O [
و
ARMA ( p, q ) AIC ( m ) min AIC ( m )
337
ARMA ( 0, 0 ) NA
ARMA (1, 0 ) 98.928
ARMA ( 2, 0 ) 52.288 *
ARMA ( 0,1) 95.912
ARMA ( 0, 2 ) 65.271
ARMA (1,1) 67.101
ARMA ( 2,1) NA
ARMA (1, 2 ) 57.818
ARMA ( 2, 2 ) 53.6
ARMA ( 2, 0 ) ذج1!'
AIC 7CاA
&ت ا1"!
!"ر ا4!B HB ان أ5
:7Bا1(
وإ;(ر اo%N -2
ACF of Residuals for Observed
(with 95% confidence limits for the autocorrelations)
1.0
0.8
0.6
Autocorrelation
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 4 6 8 10 12 14 16
Lag
338
PACF of Residuals for Observed
(with 95% confidence limits for the partial autocorrelations)
1.0
0.8
Partial Autocorrelation
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2 4 6 8 10 12 14 16
Lag
Normal Probability Plot of the Residuals
(response is Observed)
2
Residual
-1
-2
-3
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
Normal Score
339
Histogram of the Residuals
(response is Observed)
10
Frequency
-3 -2 -1 0 1 2 3 4
Residual
Norm al Probability Plot
.999
.99
.95
Probability
.80
.50
.20
.05
.01
.001
-3 -2 -1 0 1 2 3
RESI1
Average: 0.0042206 Ko lmo go rov-Smirno v No rmality Test
StDev: 1.39083 D+: 0.067 D-: 0.068 D : 0.068
N: 64 Appro ximate P-Value > 0.15
.W'& ;حJ!
ذج ا1!'
ان ا7'"2 اA;(رات وه9 اH ;ز آC 7Bا1(
أن ا5
95% :('C ;اتN V& مJC 4'& أز8 ;5 4(J; !
) اJ
ات:('C -3
340
Forecasts from period 64
95 Percent Limits
Period Forecast Lower Upper
Actual
65 197.419 194.648 200.190
66 198.729 194.059 203.400
67 200.196 194.621 205.772
68 201.215 195.480 206.949
69 201.507 195.756 207.258
70 201.148 195.181 207.116
71 200.445 194.202 206.687
72 199.756 193.378 206.134
207
202
C6
197
192
In d e x 10 20 30 40 50 60 70
341
:%nا
ا
1- Abraham, B. and Ledoter, J. (1983). Statistical Methods for
Forecasting, John Wiley, New York.
2- Anderson, T. W. (1971). The Statistical Analysis of Time Series,
John Wiley, New York.
3- Box, G. E. P. and Jenkins, G. M. (1976). Time Series Analysis
Forecasting and Control, 2nd ed., Holden-Day, San Francisco.
4- Montgomery, D. C., Johnson, L. A. and Gardiner, J. S. (1990).
Forecasting and Time Series Analysis, 2nd ed., McGraw-Hill
International Edition.
5- Makridakis, S., Wheelwright, S. C. and McGee, V. E. (1983).
Forecasting Methods and Applications, 2nd ed., John Wiley, New
York.
6- Wei, W. W. S. (1990). Time Series Analysis Univariate and
Multivariate Methods, Addison Wesley.
7- Minitab Reference Manual, Release 11 for Windows. (1998).
ة6Y!
ى ا
!دة ا1;%& & ءL أوH آ76YC 4. VL &ا7! W 5 L1CX 2K
اe
;ب أن3
او اVL!
ا اA هH]!. )"2 أو &رسQ5. أوW
= ا & أي1L;ب وأر3
ا اA ه7N
:7
;و3
9 ا2(
ا4I5 & 7
H2
abarry@ksu.edu.sa أوabarry@abarry.ws
342
You might also like
- The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good LifeFrom EverandThe Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good LifeRating: 4 out of 5 stars4/5 (5834)
- The 7 Habits of Highly Effective PeopleFrom EverandThe 7 Habits of Highly Effective PeopleRating: 4 out of 5 stars4/5 (354)
- The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good LifeFrom EverandThe Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good LifeRating: 4.5 out of 5 stars4.5/5 (20146)
- Pride and Prejudice: Bestsellers and famous BooksFrom EverandPride and Prejudice: Bestsellers and famous BooksRating: 4.5 out of 5 stars4.5/5 (20479)
- Remarkably Bright Creatures: A NovelFrom EverandRemarkably Bright Creatures: A NovelRating: 4.5 out of 5 stars4.5/5 (5943)
- Girl, Wash Your Face: Stop Believing the Lies About Who You Are so You Can Become Who You Were Meant to BeFrom EverandGirl, Wash Your Face: Stop Believing the Lies About Who You Are so You Can Become Who You Were Meant to BeRating: 4.5 out of 5 stars4.5/5 (2150)
- Never Split the Difference: Negotiating As If Your Life Depended On ItFrom EverandNever Split the Difference: Negotiating As If Your Life Depended On ItRating: 5 out of 5 stars5/5 (3334)
- Art of War: The Definitive Interpretation of Sun Tzu's Classic Book of StrategyFrom EverandArt of War: The Definitive Interpretation of Sun Tzu's Classic Book of StrategyRating: 4 out of 5 stars4/5 (3321)
- The 7 Habits of Highly Effective People: The Infographics EditionFrom EverandThe 7 Habits of Highly Effective People: The Infographics EditionRating: 4 out of 5 stars4/5 (2488)
- Habit 3 Put First Things First: The Habit of Integrity and ExecutionFrom EverandHabit 3 Put First Things First: The Habit of Integrity and ExecutionRating: 4 out of 5 stars4/5 (2510)
- American Gods: The Tenth Anniversary EditionFrom EverandAmerican Gods: The Tenth Anniversary EditionRating: 4 out of 5 stars4/5 (12968)
- The 7 Habits of Highly Effective PeopleFrom EverandThe 7 Habits of Highly Effective PeopleRating: 4 out of 5 stars4/5 (2572)
- Habit 1 Be Proactive: The Habit of ChoiceFrom EverandHabit 1 Be Proactive: The Habit of ChoiceRating: 4 out of 5 stars4/5 (2562)
- How To Win Friends And Influence PeopleFrom EverandHow To Win Friends And Influence PeopleRating: 4.5 out of 5 stars4.5/5 (6715)
- The Iliad: A New Translation by Caroline AlexanderFrom EverandThe Iliad: A New Translation by Caroline AlexanderRating: 4 out of 5 stars4/5 (5734)
- Wuthering Heights (Seasons Edition -- Winter)From EverandWuthering Heights (Seasons Edition -- Winter)Rating: 4 out of 5 stars4/5 (9976)
![American Gods [TV Tie-In]: A Novel](https://imgv2-1-f.scribdassets.com/img/audiobook_square_badge/626321117/198x198/22ab6b48b6/1718315036?v=1){kind=icon}
