C8 - A1 - Feril

You might also like

Download as xlsx, pdf, or txt
Download as xlsx, pdf, or txt
You are on page 1of 6

Consider the following time series data:

Month 1   2 3 4 5 6
Value      24 13 20 12 19 23

a. Construct a time series plot. What type of pattern exists in the data?

30 Time Series Value     


25
VALUE

20

15

10

0
1    2 3 4 5 6 7
MONTH

It denotes the occurrence of a horizontal pattern at the set time series value.
b. Develop a three-week moving average for this time series. Compute MSE and a
forecast for week 8.
Week Value 3 week AVE. Forcast Forcast Error Forcast Error^2
1 24
2 13 19
3 20 15
4 12 17 19 -7 49
5 19 18 15 4 16
6 23 19 17 6 36
7 15 18 -3 9
Total: 110

Compute 3 week moving Ave


𝑀𝐴_2=(24+13+20)/3=19 𝑀𝐴_3=(13+20+12)/3=15
Forcast Error 〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒−Forc

𝑀𝐴_4=(20+12+19)/3=17 𝑀𝐴_5=(12+19+23)/3=18 〖𝐹𝐸〗 _4=12−19=−7

𝑀𝐴_6=(19+23+15)/3=19 〖𝐹𝐸〗 _5=19−15=4

MSE = Average of the sum of squared forecast error


𝑀𝑆𝐸=(49+16+36+9)/4=27.
5
Forcast for week 8

〖𝐹𝐶〗 _8=(19+23+
15)/3=19

c. Use an alpha = 0.2 to compute the exponential smoothing values for the time series.
Compute MSE and a forecast for week 8.
Week Value Ave. smoothing Forcast Forcast Error Forcast Error^2
1 24
2 13 24 -11 121
3 20 21.8 -1.8 3.24
4 12 21.44 -9.44 89.1136
5 19 19.552 -0.552 0.304704
6 23 19.4416 3.5584 12.66221056
7 15 20.15328 -5.15328 26.5562947584
Total: 252.876809318

Equation and computation for Ave. Smoothing Forcast

〖𝐹𝐶〗 _𝑤𝑒𝑒𝑘=0.2 (𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝐴𝑐𝑡𝑢𝑎𝑙 𝑉𝑎𝑙𝑢𝑒)+(1 −0.2) (𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝐹𝑜𝑟𝑐𝑎𝑠𝑡)


𝛼=0.2

〖𝐹𝐶〗 _2=0.2 (24)+(1−0.2)(24)= 24 〖𝐹𝐶〗 _5=0.2 (12)+(1−0.2)(21.44)= 19.552

〖𝐹𝐶〗 _3=0.2 (13)+(1−0.2)(24)= 21.8 〖𝐹𝐶〗 _6=0.2 (19)+(1−0.2)(19.552)= 19.4416

〖𝐹𝐶〗 _4=0.2 (20)+(1−0.2)(21.8)= 21.44 〖𝐹𝐶〗 _7=0.2 (23)+(1−0.2)(19.5516)= 20.15328


MSE = Average of the sum of squared forecast error
𝑀𝑆𝐸=252.8768093/6=𝟒
𝟐.𝟏𝟒𝟔

d. Compare the three-week moving average forecast with the exponential smoothing forecast using an alpha = 0.2. Which a

I was able to solve for MSE = 27.5 and the Forcast for week 8 FC8 = 19 from the issue from the three week moving average,
with alpha = 0.2 from the exponential smoothing with alpha =

As a result, MSE (Moving Average) is less than MSE (Exponential Smoothing Average), with a value of 27.5 42.146. As a co

e. Use trial and error to find a value of the exponential smoothing coefficient alpha that results in a smaller MSE than what
To begin, we must first decide which alpha to test. We'll use this function (0.1 0.2 > 0.3) so we're starting with 0.2.
When 𝛼=0.1
Week Value Ave. smoothing Forcast Forcast Error Forcast Error^2
1 24 Equation and computation
2 13 24 -11 121 𝛼=0.1
3 20 22.9 -2.9 8.41
4 12 22.61 -10.61 112.5721
5 19 21.549 -2.549 6.497401
6 23 21.2941 1.7059 2.91009481 To calculate forcast error t
7 15 21.46469 -6.46469 41.7922167961 〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒
Total: 293.181812606

MSE = Average of the sum of squared forecast error


𝑀𝑆𝐸=293.1818126/6=
48.86363543

When 𝛼=0.3
Week Value Ave. smoothing Forcast Forcast Error Forcast Error^2
1 24 Equation and computation
2 13 24 -11 121 𝛼=0.3
3 20 20.7 -0.7 0.49
4 12 20.49 -8.49 72.0801
5 19 17.943 1.057 1.117249
6 23 18.2601 4.7399 22.46665201 To calculate forcast error t
7 15 19.68207 -4.68207 21.9217794849 〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒
Total : 239.075780495

MSE = Average of the sum of squared forecast error


𝑀𝑆𝐸=293.1818126/6=
39.84596342

As a result, we'll use the equation (0.1 0.2 > 0.3) to come up with this.
𝑀𝑆𝐸_(𝛼=0.1)=48.864> 𝑀𝑆𝐸_(𝛼=0.2)=42.416> 𝑀𝑆𝐸_(𝛼=0.3)=𝟑𝟗.𝟖𝟒𝟓

We can now say that achieving a minimum MSE, which is good for exponential smoothing forecasting, is foreseeable.
MSE alpha 0.3 has a lower MSE than alpha 0.2.
7
15

6 7

Forcast Error^2

〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒−Forcast

〖𝐹𝐸〗 _6=23−17=6

〖𝐹𝐸〗 _7=15−18 =−3


To calculate forcast error this is the equation

〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒−Forcast 〖〖 (𝐹
𝐸 〗 _𝑛)
−0.2)(21.44)= 19.552 〗 ^2

0.2)(19.552)= 19.4416

0.2)(19.5516)= 20.15328
Forcast for week 8

〖𝐹𝐶〗 _8=0.2 (15)+(1−0.2)


(20.15)=𝟏𝟗.𝟏𝟐

sing an alpha = 0.2. Which appears to provide the better forecast based on MSE?

hree week moving average, and MSE = 42.146 and FC8 = 19.12 from the exponential smoothing
ntial smoothing with alpha = 0.2.

value of 27.5 42.146. As a consequence, the moving average outdoes exponential smoothing.

n a smaller MSE than what you calculated for an alpha = 0.2.


tarting with 0.2.

Equation and computation for Ave. Smoothing Forcast


〖𝐹𝐶〗 _𝑤𝑒𝑒𝑘=0.1 (𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝐴𝑐𝑡𝑢𝑎𝑙 𝑉𝑎𝑙𝑢𝑒)+(1 −0.1) (𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝐹𝑜𝑟𝑐𝑎𝑠𝑡)
〖𝐹𝐶〗 _2=0.1 (24)+(1−0.1)(24)= 24
〖𝐹𝐶〗 _2=0.1 (24)+(1−0.1)(24)= 24

To calculate forcast error this is the equation


〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒−Forcast 〖〖 (𝐹𝐸 〗 _𝑛) 〗 ^2

Equation and computation for Ave. Smoothing Forcast


〖𝐹𝐶〗 _𝑤𝑒𝑒𝑘=0.3 (𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝐴𝑐𝑡𝑢𝑎𝑙 𝑉𝑎𝑙𝑢𝑒)+(1 −0.3) (𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝐹𝑜𝑟𝑐𝑎𝑠𝑡)
〖𝐹𝐶〗 _2=0.3 (24)+(1−0.3)(24)= 24

To calculate forcast error this is the equation


〖𝐹𝐸〗 _𝑛=𝑉𝑎𝑙𝑢𝑒−Forcast 〖〖 (𝐹𝐸 〗 _𝑛) 〗 ^2

asting, is foreseeable.

You might also like