FUZZY INFERENCE SYSTEM

J.Saminathan
1513419867
Department of ECE
Anna University, Chennai
 Fuzzy Inference System
 A Fuzzy Inference System (FIS) is a way of mapping an

input space to an output space using fuzzy logic

 The primary work of this system is decision-making.

 FIS uses “IF...THEN” rules along with connectors “OR”

or “AND” for making necessary decision rules.

 The input to FIS may be fuzzy or crisp, but the output

from FIS is always a fuzzy set.

The Architecture of Fuzzy Inference
Systems
Inference
Input Fuzzifier Defuzzifier Output
Engine

Fuzzy
Knowledge base
 Fuzzy Knowledge Base
I nf erence
I nput Fuzzifi er Def uzzifi er Output
Engine

Fuzzy
Knowledge base

 The rule base and the database are jointly referred

to as the knowledge base.
 a rule base containing a number of fuzzy IF–THEN rules;
 a database which defines the membership functions of
the fuzzy sets used in the fuzzy rules.
Fuzzifier
I nf erence
I nput Fuzzifi er Def uzzifi er Output
Engine

Fuzzy
Knowledge base

Converts the crisp input to a linguistic variable

using the membership functions stored in the
fuzzy knowledge base.
 I nf erence
I nput Fuzzifi er Def uzzifi er Output
Engine

Inference Engine
Fuzzy
Knowledge base

Using If-Then type fuzzy rules converts the

fuzzy input to the fuzzy output.
Defuzzifier
I nf erence
I nput Fuzzifi er Def uzzifi er Output
Engine

Fuzzy
Knowledge base

Converts the fuzzy output of the inference

engine to crisp using membership functions
similar to the ones used by the fuzzifier.
Fuzzy Inference Systems

Mamdani
Fuzzy models
 Mamdani Fuzzy models
Original Goal: Control a steam engine & boiler

combination by a set of linguistic control rules

obtained from experienced human operators.

 Mamdani fuzzy inference
 The Mamdani-style fuzzy inference process is
performed in four steps:
 A Fuzzifier which translates real - valued inputs into
fuzzy values.

 An inference engine that applies a fuzzy reasoning

mechanism to obtain a fuzzy output

 Defuzzifier, Which translates fuzzy into a crisp value

 Knowledge base, which contains fuzzy rules

The Reasoning Scheme
Max-Min Composition is used.
The Reasoning Scheme
Max-Product Composition is used.
 I nf erence

Defuzzifier
I nput Fuzzifi er Def uzzifi er Output
Engine

Fuzzy
Knowledge base

 Converts the fuzzy output of the inference engine to crisp

using membership functions similar to the ones used by
the fuzzifier.
 Commonly used defuzzifying methods:
 Centroid of area (COA)
 Weighted average (WA)
 Mean of maximum (MOM)
 Center of sums (COS)
 Center of largest area (CLA)
 First of maxima (FOM)
Mamdani Fuzzy Inference
We examine a simple two-input one-output problem that includes three
rules:

Rule: 1 Rule: 1
IF x is A3 IF project_funding is sufficient
OR y is B1 OR project_staffing is small
THEN z is C1 THEN risk is low

Rule: 2 Rule: 2
IF x is A2 IF project_funding is marginal
AND y is B2 AND project_staffing is large
THEN z is C2 THEN risk is normal

Rule: 3 Rule: 3
IF x is A1 IF project_funding is insufficient
THEN z is C3 THEN risk is high
Step 1: Fuzzification
 The first step is to take the crisp inputs, x1 and y1 (project
funding and project staffing), and determine the degree to which
these inputs belong to each of the appropriate fuzzy sets.

Crisp Input Crisp Input

x1 y1

1 1 B1 B2
A1 A2 A3 0.7
0.5
0.2 0.1
0 0
x1 X y1 Y
 
(x = A1) = 0.5 (y = B1) = 0.1
 = 0.2  = 0.7
(x = A2) (y = B2)
Step 2: Rule Evaluation
RECAL:
To evaluate the disjunction of the rule antecedents, we use the OR
fuzzy operation. Typically, fuzzy expert systems make use of the
classical fuzzy operation union:

AB(x) = max [A(x), B(x)]

Similarly, in order to evaluate the conjunction of the rule

antecedents, we apply the AND fuzzy operation intersection:

AB(x) = min [A(x), B(x)]

 Step 2: Rule Evaluation
1 1 1
A3 B1 C1 C2 C3
0.1 OR 0.1
0.0
(max)
0 x1 X 0 y1 Y 0 Z

Rule 1: IF x is A3 (0.0) OR y is B1 (0.1) THEN z is C1 (0.1)

1 1 1
0.7
C1 C2 C3
A2 0.2 B2 AND 0.2
(min)
0 x1 X 0 y1 Y 0 Z
Rule 2: IF x is A2 (0.2) AND y is B2 (0.7) THEN z is C2 (0.2)
1 1
A1 0.5 0.5 C1 C2 C3

0 x1 X 0 Z
Rule 3: IF x is A1 (0.5) THEN z is C3 (0.5)
 Step 3: Aggregation of the rule outputs

1 1 1
C1 C2 C3
0.5 0.5
0.2 0.2
0.1 0.1
0 Z 0 Z 0 Z 0 Z
z is C 1 (0.1) z is C 2 (0.2) z is C 3 (0.5) 
 Step 4: Defuzzification
Degree of
Membership
1.0

0.8

0.6

0.4

0.2

0.0
0 10 20 30 40 50 60 70 80 90 100
67.4 Z

(0  10  20) 0.1  (30  40  50  60) 0.2  (70  80  90  100) 0.5

COA  67.4
0.1  0.1  0.1  0.2  0.2  0.2  0.2  0.5  0.5  0.5  0.5
Fuzzy Inference Systems

Sugeno
Fuzzy Models
 Sugeno Fuzzy Inference
 In Mamdani FIS, it requires us to find the centroid of a two-

dimensional shape by integrating across a continuously

varying function.
 In general, this process is computationally not efficient

one.
 Sugeno suggested to use a single spike or a singleton, as

the membership function of the rule consequent.

 Also known as TSK fuzzy model
 Takagi, Sugeno & Kang, 1985
 Sugeno Fuzzy Inference
 Sugeno-style fuzzy inference is very similar to the Mamdani method.

Sugeno changed only a rule consequent.

 Instead of a fuzzy set, he used a mathematical function of the

input variable.
 The format of the Sugeno-style fuzzy rule is

IF x is A

AND y is B

THEN z is f(x, y)

where x, y and z are linguistic variables; A and B are fuzzy sets on

universe of discourses X and Y, respectively; and f(x, y) is a
mathematical function.
 Sugeno Fuzzy Inference
 The most commonly used zero-order Sugeno fuzzy model

applies fuzzy rules in the following form:

IF x is A

AND y is B

THEN z is k

where k is a constant.
 In this case, the output of each fuzzy rule is constant.

 All consequent membership functions are represented by

singleton spikes.
Sugeno Rule Evaluation
1 1 1
A3 B1
0.1 OR 0.1
0.0
(max)
0 x1 X 0 y1 Y 0 k1 Z

Rule 1: IF x is A3 (0.0) OR y is B1 (0.1) THEN z is k1 (0.1)

1 1 1
0.7
A2 0.2 B2 AND 0.2
(min)
0 x1 X 0 y1 Y 0 k2 Z
Rule 2: IF x is A2 (0.2) AND y is B2 (0.7) THEN z is k2 (0.2)
1 1
A1 0.5 0.5

0 x1 X 0 k3 Z
Rule 3: IF x is A1 (0.5) THEN z is k3 (0.5)
 Sugeno Aggregation of the Rule
Outputs

1 1 1 1
0.5 0.5
0.2
0.1 0.2 0.1
0 k1 Z 0 k2 Z 0 k3 Z 0 k1 k2 k3 Z

z is k1 (0.1) z is k2 (0.2) z is k3 (0.5) 

 Sugeno Defuzzification
Weighted Average (WA)

(k1) k1  (k 2) k 2  (k 3) k 3 0.120  0.2 50  0.5 80

WA   65
(k1)  (k 2)  (k 3) 0.1  0.2  0.5

0 z1 Z

Crisp Output
z1
Fuzzy Inference Systems

Tsukamoto
Fuzzy models
 Tsukamoto Fuzzy models
• The consequent of each fuzzy if-then-rule is
represented by a fuzzy set with a
monotonical MF.
• Crisp output is weighted average of fuzzy
outputs
Tsukamoto fuzzy model
 Case Study

Comparison of Mamdani and Sugeno Fuzzy

Interference Systems for the Breast Cancer
Risk
 Breast Cancer
 Breast cancer is a major health burden worldwide.

 Despite the billions of dollars spent on breast cancer

research, incidence rates have been climbing steadily in

industrialized countries since the 1940s
 Moreover, clinical Oncologists make diagnostic decisions

about breast cancer patients based on past professional

experience, knowledge, and intelligent techniques
 Due to their stable behaviour in the presence of noise,

imprecision and uncertainty, FIS techniques could potentially

obtain better results than classical methods
 FUZZY INFERENCE SYSTEM
 Fuzzy inference is the process of formulating the mapping
from given input(s) to output(s) using fuzzy logic.
 This mapping provides a basis from which decisions can be
made
 A rule base containing a number of fuzzy if-then rules.
 A database which defines the membership functions of the
fuzzy sets used in fuzzy rules.
 A decision-making unit which performs the inference
operations on the rules.
 A fuzzification interface which transforms the crisp inputs
into degrees of match with linguistic values.
 A defuzzification interface which transform the fuzzy results
of the inference into a crisp output.
For the evaluation of breast cancer risk the
proposed FIS consists of two inputs: age and tumor
surface.
The system has one output that indicates the risk of
breast cancer in percentage
The age and the tumor surface (i.e., tumor size) of
patients are selected as input variables for the fuzzy
system due to the most relevance factors decided
by clinicians
The breast cancer risk factor is the output of the FIS
 The age and the tumor size are taken to be in ranges of 0 to

80 years and 0 to 8000 pixels, respectively.

 Each of the selected input and output variables is described

by a set of three linguistic fuzzy values, defined by a

Gaussian membership function, thus allowing the

fuzzification procedure to convert the measured numerical

value into one of the fuzzy values.

 Age MF

Tumor Size MF

Breast Cancer Risk MF

By using the graphical user interface of the Fuzzy
Logic Toolbox, the fuzzy inference system for the
Mamdani-type
• Unlike the output value range of the Mamdani-type fuzzy inference
system, the range of Sugeno-type output is between 0 and 1.
• The output of this system can only be either constant or linear in this FIS,
so three linguistic fuzzy values for the output are “Low”, “Medium”, and
“High” which can be constant
 Conclusion
It can be concluded that Mamdani-type FIS and
Sugeno-type FIS perform quite similar, but Sugeno-type
FIS allows the evaluation of risk to work at its full
capacity with smooth operational performance.
 FIS: Sugeno vs. Mamdani
 Advantages of the Sugeno Method
 It is computationally efficient.
 It can be used to model any inference system in which the
output membership functions are either linear or constant.
 It works well with optimization and adaptive techniques.
 It is well suited to mathematical analysis.
 Advantages of the Mamdani Method
 It is natural.
 It has widespread acceptance.
 It is well suited to human input
Reference
Alshalaa A. Shleeg, Issmail M. Ellabib, (2013),
“Comparison of Mamdani and Sugeno Fuzzy
Interference Systems for the Breast Cancer Risk”,
International Journal of Computer, Electrical,
Automation, Control and Information Engineering
Vol:7, No:10, pp. 1343-1347.
Thank You