KECERDASAN BERBASIS

PENGETAHUAN

INFERENSI DALAM
KETIDAK PASTIAN
KECERDASAN BERBASIS
PENGETAHUAN

CERTAINTY FACTOR
Uncertainty: Introduction

 Dalam Sistem berbasis pengetahuan, sering kali

sistem harus mengambil kesimpulan yang benar
dari bukti yang lemah dan tidak pasti

 Bukan hal yang mustahil. Sering kita lakukan

dengan sukses dalam hampir semua aspek
kehidupan kita untuk dapat bertahan hidup.

3
Uncertainty: Introduction

 Misal :
- kita mengerti suatu kalimat yang tidak lengkap
- Dokter memberikan medical treatment yang
benar untuk symptom yang ambigu;

 Salah satu karakteristik dari informasi yang tersedia bagi

seorang pakar adalah ketidaksempurnaan, yaitu informasi
dapat : incomplete, inconsistent, uncertain

4
So, how to define the term “Uncertainty”?

 Uncertainty didefinisikan sebagai : kurangnya ”exact

knowledge” yang memungkinkan kita membuat
konklusi yang tepat.
 Hal ini disebabkan karena informasi yang tersedia
tidak sempurna, seperti : tidak konsisten, tidak
lengkap, tidak pasti
 Can be a problem because it may prevent us from
making the best decision or may even cause bad
decision to be made

5
So, how to define the term “Uncertainty”?

 Reasoning under uncertainty involves 3 important

issues:
 How to represent uncertain data
 How to combine two or more pieces of uncertain
data
 How to draw inference using uncertain data
There are many approaches to
representing uncertainty in AI.

7
Uncertainty Handling Methods
 Abductive reasoning
 Property inheritance Abductive reasoning would say that
 Fuzzy logic
 Certainty Factor (CF)
 Bayes theorem
 Dempster-Shafer theory
E.g. Suppose:
If x is a bird then x flies
“All fly things are birds”
By property inheritance
“All birds can fly”
but, remember the case that
Penguin cannot fly?
8
 Evaluation Criteria for
uncertainty handling methods

 Expressive power
 Logical correctness
 Computational efficiency of
inference

9
 Scheme used by expert system in
Handling Uncertainty

 MYCIN (sistem pakar deteksi penyakit kelainan

darah ) uses Certainty Factor
 The CF can be used to rank hypotheses in order of
importance.
 For example if a patient has certain symptoms that suggest
several possible diseases, then the disease with the higher
CF would be the one that is first investigated.
 PROSPECTOR (sistem pakar untuk eksplorasi
mineral) uses Bayes theorem

10
UNCERTAIN EVIDENCE
 Representing uncertain evidence
CF (E) = CF (It will probably rain today)
= 0.6
 Similar to P(E), however, CF are not probabilities
but measures of confidence
 Derajat yang menunjukkan kita percaya bahwa
fakta benar
 It written as exact term but add CF value to it
Example: It will rain today CF 0.6
 Certainty Theory
 Certainty factors measure the confidence that is
placed on a conclusion based on the evidence
known so far.
 Certainty factors combine belief and disbelief
into a single number based on some evidence
 A certainty factor is the difference between the
following two components :
CF = MB[h:e] - MD[h:e]

A positive CF means the evidence supports the hypothesis since MB > 12

MD.
Certainty Theory

1. Measures of Belief (MB)

Number that reflects the measure of increased belief
in a hypothesis H based on evidence E
0  MB  1

2. Measures of Disbelief (MD)

Number that reflects the measure of increase
disbelief in a hypothesis H based on evidence E
0  MD  1
Certainty Theory

3. Certainty Factor
Number that reflects the net level of belief in a
hypothesis given available information
CF = MB - MD
-1  CF 1
Certainty Theory

 Uncertain evidence is given CF or certainty factor value

ranging from -1 (completely false ) to 1 (completely true).

 Negative values degree of disbelief

 Positive values degree of belief

 Range of CF values

false Possibly False Unknown Possible True True

-1 0 1
Measures of disbelief Measures of bilief
Certainty Theory: Values Interpretation

Definitely Not -1.0

Almost certainly not -0.8
Probably not -0.6
Maybe not -0.4
Unknown -0.2 to 0.2
Maybe 0.4
Probably 0.6
Almost Certainly 0.8
Definitely 1.0
Certainty Theory
 Basic structure of rule in certainty model
IF E THEN H CF (Rule)

 CF (Rule) = level of belief of H given E

 Given that E is true, we believe H according to:

CF (H, E) = CF (Rule)
Certainty Theory

 Example:
IF There are dark cloud -E
THEN It will rain - H CF = 0.8

 This rule reads:

“If there are dark clouds then it will almost
certainly rain”

 CF (H, E) similar P (H|E)

 Certainty Factor Propagation (C.F.P)

CFP FOR SINGLE PREMISE RULE

Menetapkan tingkat kepercayaan pada konklusi suatu rule (H)
ketika fakta (E) yang termuat dalam rule premise tidak pasti

CF (H, E) = CF (E) X CF (RULE)

Example
From previous rule:
CF (E) = 0.5
C.F.P. is
CF (H, E) = 0.5 x 0.8
= 0.4 #
In words: It maybe raining
Certainty Factor Propagation (C.F.P)

For same rule, with negative evidence for rule premise

to the degree of CF (E) = -0.5

C.F.P is
CF (H, E) = -0.5 * 0.8
= -0.4 #

In words: It maybe not raining.

Certainty Factor Propagation (C.F.P)

CFP FOR MULTIPLE RULE PREMISE

 For rules with more than one premises:

– conjunctive rules (AND)
– disjunctive rules (OR)
Certainty Factor Propagation (C.F.P)

CFP FOR MULTIPLE RULE PREMISE

 Conjunctive Rule
IF EI AND E2 AND ... THEN H CF (RULE)
CF (H, E1 AND E2 AND..) = min {CF (Ei)} * CF
(RULE)

 min function returns minimum value of a set of

numbers.
Certainty Factor Propagation (C.F.P)

Example:
IF The sky is dark
AND The wind is getting stronger
THEN It will rain CF = 0.8
Assume:
CF (the sky is dark) = 1.0
CF (the wind is getting stronger) = 0.7
CF (It will rain) = min {1.0,0.7} * 0.8
= 0.56 #
In words: It probably will rain
Certainty Factor Propagation (C.F.P)

CFP FOR MULTIPLE RULE PREMISE

 Disjunctive Rule (OR)

IF E1 OR E2 OR ... THEN H CF (RULE)

CF (H, E1 OR E2 OR..) = Max {CF (Ei)} * CF
(RULE)

 Max function returns the maximum value of a set

of numbers
Certainty Factor Propagation (C.F.P)

 Example:
IF the sky is dark
OR the wind is getting stronger
THEN It will rain CF = 0.8
CF (It will rain) = max {1.0, 0.7} * 0.8
= 0.8 #

 In words: It almost certainly will rain

Certainty Factor Propagation (C.F.P)

CFP SIMILARLY CONCLUDED RULES

 For multiple rules that support a
hypothesis (same hypothesis)

 Consider from 2 individuals:

– weatherman
– farmer
Certainty Factor Propagation (C.F.P)

Rule 1
IF the weatherman says it is going to rain (E1)
THEN It is going to rain (H)
CF (Rule 1) = 0.8

Rule 2
IF the farmer says it is going to rain
(E2)
THEN It is going to rain (H)
CF (Rule 2) = 0.8
Certainty Factor Propagation (C.F.P)

 CF of both rules set to equal implying equal

confidence in the 2 sources

 Naturally more confident in conclusion

Certainty Factor Propagation (C.F.P)

FORMULA FOR INCREMENTALLY ACQUIRED EVIDENCE

CFcombined (CF1, CF2)

= CF1 + CF2 * (1 - CF1) Both  0

= CF1 + CF2 One  0

1 - min {|CF1|,|CF2|}

= CF1 + CF2 * (1 + CF1) Both  0

where,
CF1 = confidence in H established by one rule (RULE 1)
CF2 = confidence in it established by one rule (RULE 2)
CF1 = CF1 (H, E)
CF2 = CF2 (H, E)
Certainty Factor Propagation (C.F.P)

 There are 2 properties of the equations:

– Commutative:
 allow evidence to be gathered in any order
– Asymptotic:
 If more than one source confirm a
hypothesis then a person will feels more
confident. Incrementally add belief to a
hypothesis as new positive evidence is
obtained
Certainty Factor Propagation (C.F.P)

Example
Consider Rain Prediction: Rule 1 and 2
Explore several cases

Case 1: Weatherman and Farmer Certain in Rain

CF (E1) = CF (E2) = 1.0

CF1 (H, E1) = CF (E1) * CF (RULE 1)

C.F.P.
= 1.0 * 0.8 = 0.8 for
single
CF2 (H, E2) =CF (E2) * CF (RULE 2) Premise
Rule
= 1. 0 * 0.8 = 0.8
Certainty Factor Propagation (C.F.P)

Since both > 0,

CFcombine (CF1, CF2) = CF1 + CF2 * (1 - CF1)

= 0.8 + 0.8 * (1 - 0.8)
= 0.96 #

CF supported by > 1 rule can be incrementally

increase more confident
Certainty Factor Propagation (C.F.P)

Case 2: Weatherman certain in rain, Farmer certain no rain

CF (E1) = 1.0 CF (E2) = -1.0
CF1 (H, E1) = CF (E1) * CF (RULE 1)
= 1.0 * 0.8 = 0.8

CF2 (H, E2) = CF (E2) * CF (RULE 2)

= -1.0 * 0.8 = -0.8
Since either one < 0
CFcombined (CF1, CF2) = CF1 + CF2
1 - min {|CF1|,|CF2|}
= 0.8 + (-0.8)
1 - min {0.8,0.8}
=0#
CF set to unknown because one say “no” and the other one say “yes”
 Certainty Factor Propagation (C.F.P)

Case 3: Weatherman and Farmer believe at different degrees that it

is going to rain
CF (E1) = -0.8 CF (E2) = -0.6
CF1 (H, E1) = CF (E1) * CF (RULE 1)
= -0.8 x 0.8
= -0.64
CF2 (H, E2) = CF (E2) * CF (RULE 2)
= -0.6 * 0.8
= -0.48
Since both < 0
CFcombined (CF1, CF2) = CF1 + CF2 * (1 + CF1)
= - 0.64 - 0.48 * (1 - 0.64)
= -0.81 #
Show incremental decrease when more than one source
disconfirming evidence is found
 Certainty Factor Propagation (C.F.P)

Case 4: Several Sources Predict Rain at the same level of belief but one
source predicts no rain
If many sources predict rain at the same level, CF(Rain) = 0.8, the CF
value converge towards 1
CFcombined (CF1, CF2 ..) 0.999 = CFold
CFold = collected old sources info.
If new source say negative
CFnew = -0.8 then,
CFcombined (CFold, CFnew)
= CFold + CFnew
1 - min { CFold, CFnew }
= 0.999-0.8
1 - 0.8
= 0.995
Shows single disconfirming evidence does not have major impact
Certainty Factor Propagation (C.F.P)

CERTAINTY PROPAGATION FOR COMPLEX RULES

Combination of conjunctive and disjunctive statement
(“AND”, “OR”)

Example:
IF E1
AND E2
OR E3 max
AND E4
THEN H

CF (H;E)= Max {min (E1, E2), min (E3, E4)} * CF (RULE)

 Certainty factors
 Consider two rules:
(R1) hasHair  mammal CF(R1) = 0.9
(R2) forwardEyes & sharpTeeth  mammal CF(R2) = 0.7
 We now know that:
 CF(hasHair) = 0.8 CF(forwardEyes) = 0.75 CF(sharpTeeth) = 0.3
 CF(forwardEyes  sharpTeeth) = min( 0.75, 0.3 ) = 0.3

Given the premise CF, how do you combine with the CF for the rule?
CF(H, Rule) = CF(E) * CF(Rule)
So, CF(mammal, R1) = CF(hasHair) * CF(R1) = 0.8 * 0.9 = 0.72
CF(mammal, R2) = CF(forwardEyes  sharpTeeth) * CF(R2)
= 0.3 * 0.7
= 0.21
Certainty factors

Given diff rules with same conclusion, how do you combine CF's?
CF(H, Rule1 & Rule2) = CF(H, Rule1) + CF(H, Rule2)*(1-
CF(H,Rule1))

So, CF(mammal, R1 & R2)

= CF(mammal, R1) + CF(mammal, R2)*(1-
CF(mammal,R1))
= 0.72 + 0.21*0.28
= 0.72 + 0.0588
= 0.7788

note: CF(mammal, R1 & R2) = CF(mammal, R2 & R1)

 LATIHAN

 To Illustrate CF through a set of rules, we consider a small problem

of deciding whether or not I should go to a ball game.

 Assume the hypothesis “I should not go to the ball game”.

Rule 1
IF the weather looks lousy E1
OR I am in a lousy mood E2
THEN I shouldn’t go to the ball game H1 CF = 0.9
LATIHAN

Rule 2
IF I believe it is going to rain E3
THEN the weather looks lousy E1 CF = 0.8

Rule 3
IF I believe it is going to rain E3
AND the weatherman says it is going to rain E4
THEN I am in a lousy mood E2 CF = 0.9
 LATIHAN

Rule 4
IF the weatherman says it is going to rain E4
THEN the weather looks lousy E1 CF = 0.7

Rule 5
IF the weather looks lousy E1
THEN I am in a lousy mood E2 CF = 0.95
LATIHAN

Assume the following CF:

I believe it is going to rain
CF(E3) = 0.95
Weatherman believes it is going to rain
CF(E4) = 0.85
Kelebihan Metode CF

 Suatu model komputasi yang sederhana

yang memungkinkan pakar mengestimasi
kepercayaaannya dalam konklusi
 Mengijinkan ekspresi kepercayaan dan
ketidakpercayaan dalam tiap hipotesis
 Menggabungkan nilai CF lebih mudah
dibandingkan metode lain
 KECERDASAN
BERBASIS PENGETAHUAN

BAYES REASONING
 TEORI PROBABILITAS

 Suatu pendekatan matematis untuk memproses

uncertain information
 probabilitas p(xi) adalah kemungkinan suatu kejadian
xiS terjadi
– Mengukur keyakinan seseorang bahwa suatu kejadian
terjadi
– non-negative values in [0,1]
– total probability of the sample space is 1
 p(xi , xiS) = 1
 RUANG SAMPEL (S)

 himpunan semua hasil yang mungkin muncul atau

terjadi pada suatu percobaan statistik / himpunan
semua event yang mungkin

Contoh :
- pelemparan mata uang  S={M, B}
- pelemparan dadu  S = {1,2,3,4,5,6}

 What is the sample space for flipping a coin 3

times?
PENDEKATAN PROBABILITAS

– Objektif Probability
 Classical probability
- asumsi : keluaran dari suatu eksperimen
mempunyai kemungkinan sama
 Experimental probability

- berdasar pada frekuensi terjadinya suatu

kejadian
– Subjective probability (CF Theories, Dempster-
Shafer, ...)
 Berdasar pada penilaian ekspert
 Contoh

– Coin: What is the probability of obtaining heads when

flipping a coin?
– A single dice: What is the probability I will roll a four?
– A jar of 30 red and 40 green jelly beans: What is the
probability I will randomly select a red jelly bean?
– Computer: In the past 20 times I used my computer, it
crashed 4 times and didn’t crash 16 times. What is the
probability my computer will crash next time I use it?
 SIFAT PROBABILITAS

 Sifat probabilitas kejadian A :

a. 0  P(A)  1

b. P(S) = 1
P (rolling a 1 or 2 or 3 or 4 or 5 or 6) = 1

c. Probabilitas kejadian A atau B (joint probability)=

P (AB) = P(A) + P(B) – P(AB)
 SIFAT PROBABILITAS (2)

d. Untuk 2 kejadian yang mutually exclusive (tidak ada 2

kejadian yang terjadi secara bersamaan pada waktu
yang sama) :
P (AB) = P(A) + P(B)

P (rolling a 1 or a 6) = P (rolling a 1) + P (rolling a 6)

= 1/6 + 1/6 = 2/6 or 1/3

e. P(A’) = 1-P(A)
P (not rolling a 5) = 1 – 1/6 = 5/6
 SIFAT PROBABILITAS (3)

. Jika ada 2 kejadian A dan B independen (keluaran kejadian

f
1 tidak berpengaruh pada keluaran kejadian berikutnya)
P(AB) = P(A).P(B)

P (rolling a 1 and then a 6) = P(rolling a 1) * Pr(rolling a 6)

= (1/6)(1/6) = 1/36

g. Jika 2 kejadian tidak independen

P(AB) = P(A).P(B|A)
CONDITIONAL PROBABILITY

– Probabilitas bersyarat : event A given that event B

has already occurred
P( A  B)
P( A / B)  , P( B)  0
P( B)
 Probabilitas Bersyarat Dalam Data
# Cuaca Temperatur Kecepatan Angin Berolah-raga
1 Cerah Normal Pelan Ya
2 Cerah Normal Pelan Ya
3 Hujan Tinggi Pelan Tidak
4 Cerah Normal Kencang Ya
5 Hujan Tinggi Kencang Tidak
6 Cerah Normal Pelan Ya

P(cuaca=cerah|olahraga = ya) = ?

Banyaknya data berolah-raga=ya adalah 4 dari 6 data maka :

P(Olahraga=Ya) = 4/6
Banyaknya data cuaca=cerah dan berolah-raga=ya adalah 4 dari 6 data maka:
P(cuaca=cerah dan Olahraga=Ya) = 4/6
4/6
P (cuaca  cerah | olahraga  ya)  1
4/6
 Probabilitas Bersyarat Dalam Data
# Cuaca Temperatur Berolahraga
1 cerah normal ya
2 cerah tinggi ya
3 hujan tinggi tidak
4 cerah tinggi tidak
5 hujan normal tidak
6 cerah normal ya

P(cuaca=cerah, temperatur=normal | olahraga=ya)= ?

Banyaknya data berolah-raga=ya adalah 3 dari 6 data maka:

P(Olahraga=Ya) = 3/6
Banyaknya data cuaca=cerah, temperatur=normal dan berolah-
raga=ya adalah 2 dari 6 data maka :
P(cuaca=cerah, temperatur=normal, Olahraga=Ya) = 2/6
2/6 2
P (cuaca  cerah, temperatur  normal | olahraga  ya)  
3/ 6 3
 CONDITIONAL PROBABILITY
P( A  B)
P( A | B) 
P( B)
P( B  A)
P( B | A) 
P( A)

P ( B  A)  P( B | A).P ( A)
P ( A  B )  P ( B | A).P ( A)
P ( B | A).P ( A)
P( A | B)  Bayesian Rule
P( B)
CONDITIONAL PROBABILITY

Jika kejadian B tergantung hanya pada 2 kejadian

yang mutually exclusive, yaitu A dan not A :
P( B )  P( B | A). P( A)  P( B | A). P(A)

P ( B | A).P ( A)
P( A | B) 
P ( B | A).P ( A)  P ( B | A).P (A)
BAYESIAN REASONING

 derive the probability of a cause given a symptom

 has gained importance recently due to advances in
efficiency
– more computational power available
– better methods
 especially useful in diagnostic systems
– medicine, computer help systems
 Misal rule dalam basis pengetahuan dinyatakan dalam
bentuk :
If E then H [dengan prob. p] yang berarti jika E
terjadi, maka probabiltas H akan terjadi adalah p.
BAYESIAN REASONING
Bayes’ Rule for Single Event
P ( E | H ). P ( H )
P( H | E ) 
P( E )
P ( E | H ).P ( H )
P( H | E ) 
P( E | H ).P ( H )  P ( E | H ).P (H )
P(H) = prior probability dari hipotesis H benar
P(E|H) = probabilitas hipotesis H benar akan mengakibatkan evidence
E
P(not H) = prior probability hipotesis H salah
P(E|not H) = probabilitas menemukan evidence E jika hipotesis H
salah
Prior probability dan conditional probability P(E|H) disediakan oleh
ekspert. User menyediakan evidence yang diamati dan sistem pakar
menghitung posterior probability P(H|E)
CONTOH

– Diketahui rule :
meningitis  sakit leher [prob= 0,8]
diketahui :
prob. seseorang menderita meningitis= 0,0001
prob. seseorang sakit leher = 0,1
Tentukan probabilitas seseorang kena meningitis jika
sakit leher.
 Misal M = meningitis, S = sakit leher maka:
P(M|S) = P(S|M) x P(M) / P(S)
= 0.8 × 0.0001 / 0.1 = 0.0008
BAYESIAN REASONING

 Single evidence and multiple hypoteses

P ( E | H i ). P ( H i )
P( H i | E )  m

 P( E | H
k 1
k ). P ( H k )
CONTOH

 Ani mengalami gejala ada bintik-bintik di wajahnya.

- prob. muncul bintik-bintik di wajah jika terkena cacar= 0,8
- prob. Ani terkena cacar tanpa memandang gejala
apapun=0,4
- prob. muncul bintik-bintik di wajah jika alergi = 0,3
- prob. Ani kena alergi tanpa memandang gejala apapun=
0,7
- prob. muncul bintik-bintik di wajah jika jerawatan = 0,9
- probabilitas Ani kena jerawat tanpa memandang gejala
apapun = 0,5
 Penyakit apa yang diderita Ani ?
 BAYESIAN REASONING
 Multiple Evidence and multiple hypoteses

P( E1 E2 ...En | H i ).P( H i )
P( H i | E1 E2 ...En )  m

 P( E E ...E
k 1
1 2 n | H k ).P( H k )

P( E1 | H i ).P( E2 | H i ).....P( En | H i ).P( H i )

P( H i |E 1E2 ...En )  m

 P( E | H
k 1
1 k ).P( E2 | H k ).....P( En | H k ).P( H k )
 CONTOH

 Diberikan 3 conditionally independent evidence E1, E2 dan E3

yang membentuk 3 hipotesis H1, H2 dan H3 yang mutually
exclusive. Diberikan juga nilai probabilitas prior dan
probabilitas kondisional oleh ekspert seperti pada tabel :

Hipotesis
Probabilitas i=1 i=2 i=3
P(Hi) 0,40 0,35 0,25
P(E1|Hi) 0,3 0,8 0,5
P(E2|Hi) 0,9 0,0 0,7
P(E3|Hi) 0,6 0,7 0,9
 CONTOH

 Misal evidence yang pertama diamati adalah E3,

maka : P( E3 | H1 ).P( H1 )
P ( H 1 | E3 )  0, 6.0, 40
 0,6.0,400,7.0,350,9.0,25  0,34
P( E3 | H1 ).P( H1 )  P( E3 | H 2 ).P( H 2 )  P( E3 | H 3 ).P( H 3 )

 Dengan cara yang sama :

P(H2|E3)= 0,34
P(H3|E3)= 0,32
 CONTOH
 Misal sekarang evidence E1 diamati :

P ( E1 | H1 ). P ( E3 | H1 ). P ( H1 ) 0,3.0,6.0,40
P ( H1 |E 1E3 )    0,19
m
0,3.0,6.0,40  0,8.0,7.0,35  0,5.0,9.0,25
 P( E
k 1
1 | H k ). P ( E3 | H k ). P ( H k )

 P(H2|E1 E3) = 0,52

 P(H3|E1 E3)= 0,29
CONTOH

 Misal E2 diamati, maka :

 P(H1|E1 E2E3)=0,45
 P(H2|E1 E2E3 )=0
 P(H3|E1 E2E3 )=0,55
Advantages:

- Most significant is their sound theoretical foundation in

probability theory.
- Most mature uncertainty reasoning methods
- Well defined semantics for decision making

Main disadvantage:

- They require a significant amount of probability data

to construct a knowledge base.

67