Unit III (Fuzzy Logic 1 INTRODUCTION)

Lecture 1 # Basic Concept of Fuzzy Logic

NOTE: It is an overview of both Unit III and Unit IV. You May Skip This
Lecture.
Although probability theory is an old and effective tool to handle uncertainty but it can be
applied only to the situations whose characteristics are based on the random processes, that is,
process in which the occurrence of events is strictly determined by chance. . However, in reality
there are large set of problems whose characteristics are determined by non-random processes.
Here uncertainty may arise due to partial information about the problem or information which is
not fully reliable, or inherent imprecision in the language which the problem is defined. Or due
to receipt of information from more than one source about the problem which is conflicting.
In such situations, fuzzy set theory exhibits immense potential for effective solving of the
uncertainty in the problems. Fuzziness means ‘vagueness ‘. It is an excellent mathematical tool
to handle uncertainty arises due to vagueness. Understanding human speech and recognizing
handwritten characters are some common stances where fuzziness manifests.
It was proposed by Lotfi. A. Zadeh in 1965. The fuzzy set theory provides a mechanism
for representing linguistic constructs such as “tall”, “small”, “high”, “low”, “medium” etc. In
general fuzzy logic provides an inference structure that enables appropriate human reasoning
capabilities; the theory fuzzy logic is based on related graded membership as so are the function
of cognitive process. The utility of fuzzy sets lies in their ability to model uncertain or
ambiguous data as shown in figure 3a.

Fuzzy Logic
Imprecise Decision
System
And vague data

Figure 3a a fuzzy logic system accepting imprecise data and providing a decision
In fuzzy systems, values are indicated by a number called truth values ranging from 0 to 1, where
0.0 absolute falseness 1.0 represents absolute truth.
Fuzzy sets that represent fuzzy logic provide mean to model uncertainty associated with
vagueness, imprecision and lack of information regarding a problem. Consider meaning of “short
person”. For an individual X, a short person may be one whose height below 4.25 inches. For
other individual Y, a “short person” may be whose height is below or equal to 3.90 inches. The
“short” is called linguistic descriptor. The term “short“provides some meaning to both individual
X and Y but they do not provide definition. The term short is called linguistic variable. It
represent imprecision in existing system.
Dr Zadeh proposed the set membership idea to make suitable decisions when uncertainty occurs.
Consider the “short” example discussed previously. If we take “short” as height equal to or less
that 4 feet then 3.90” would easily become the member of set “short” and 4.25” will not be
member of set “short”. The membership value is “1” if it belongs to the set and “0” if it is not a
member of set. Thus membership in the set is found to be binary, that is either the element is
member of set or not. It can be indicated as

(x) = {

Where (x) is the membership of element x in the set A and A is the entire set on the universe.
If it is said that the height is 5.6” one might think a bit before deciding whether to consider it as
short or not short (i. e. tall). Moreover one might reckon it as short for man but tall for women.
Let’s take a statement “John is short” and give it a truth value of 0.70. If 0.70 represented a
probability value it would be read as “There is 70% chance that John is short “, meaning that it is
still believed that John is either short or not, and there exist 70% chance which group he belongs
to. Fuzzy terminology actually translates to “ John’s degree of membership in the set of short
people is 0.07” by which it meant if all the people fuzzy set “short” are considered and lined up,
John positioned 70% of the way to shortest.
Fuzzy logic operates on the concept of membership. For example, the statement “Elizabeth is
old” can be translated as Elizabeth is a member of the set of old people and can be written µ
, µ is membership function that can return a value between 0 and 1 depending on the
degree of membership.
Consider figure 3b the objective term “tall” has been assigned fuzzy values. At 150 cm and
below, a person does not belong to the fuzzy class while for above 180, the person certainly
belongs to the class “tall”. The fuzzy concept “tallness “ can be extended into “short”, “medium”
and “tall” as shown in figure 3 c.

1 Tall
Membership

150 180 210 ht (cm)

Figure 3 c Graph for membership functions for fuzzy set “tall”

Short Medium Tall

1
Membership

150 180 210 Height (cm)

Figure 4 c Graph showing membership functions for fuzzy sets “short”, “medium” and
“tall”
Fuzzy logic also consists of fuzzy inference engine or fuzzy rule –base to perform approximate
reasoning somewhat similar to human brain. Computing with words seems to be slightly
futuristic phrase today since only certain aspects. of natural language can be represented by
calculus of fuzzy sets.
Fuzzy sets form the building block for fuzzy IF-THEN rules which has the general form “IF x is
̂ THEN y is ̃ “where ̂ ̃ are fuzzy sets.
Lecture # 2 Fuzzy Sets and Crisp Sets
Fuzzy vs. CRISP Sets

The logic which demands for binary (1, 0) handling is termed as CRISP in domain for fuzzy set
theory.
EX.
“Is today Sunday” Answer will be (yes/no) or (1/0)
The situation where values are accepted between “yes” and “no” or between 1 -0 is termed as
fuzzy.
Extremely Honest (1)

Very Honest (0.8)

Ex. Is Ra esh ho est Fuzzy

Honest (0.5)

Partially Honest (0.3)

Dishonest (0)

Figure 2a fuzzy membership values

The Boolean logic is based on CRISP set theory similarly fuzzy logic is based on fuzzy set
Theory.

CRISP Set- (Review of Claasical Set Theory Not so Important)

Universe of discourse or universal set is the set which with reference to a particular context
contains all possible elements having same characteristics and from which set can be formed .

For example universal set of days

Set of Days = {Sun, Mon, Tue, Wed, Thu, Fri, Sat}
Set
It is collection of well defined objects. In set a particular element must be distinguished from
other elements of set
Ex. Set of Days = {Sun, Mon Thu, Fri, Sat}
or A={x: p(x)} here x is an element of set A and satisfies property p(x)

Terminologies
Membership Indicated by . x A means element x is member of set A.
Cardinality
The no of elements in a set is called cardinality of set. Suppose a given set A= {4, 5, 6, 7} then
cardinality of set A, denoted by |A| and will be 4.
Null Set
A set with no elements is called null set and denoted by ф.
Singleton Set
It has a single element.
Subset
Given sets A, B over universal set, A is said to be subset of B is A is contained in B mean A B
(Proper subset) and A B (improper subset).
Superset
If A is subset of B, it means B is superset of A. It is represented as A

Power Set.
All possible subset of a set including null set is called the power set.

For Example if A= {2, 4, 6, 7} then power set of A, denoted by Ap is given as

{ ф, {2}, {4}, {6}, {7}, {2, 4}, {2, 6},{2,7},[4,6},(4,7},{6,7}, {2,4,6}, {2, 4, 7},{4,6,7}, {2, 6,7},
{2,4,6,7}}
Operations on sets

We can perform following operations on sets.

1. Union
Given two sets A and B, union of A and B is denoted by A B.
Given A= {1, 2, 3} and B= {2, 3, 5}
Then A B= {1, 2, 3, 5}
2. Intersection
. Given two sets A and B, intersection of A and B is denoted by A B.
Given A= {1, 2, 3} and B= {2, 3, 5}
Then A B= {2, 3}
3. Complement
Given universe of discourse U and set A, then complement of A is denoted by Ac . If U=
{1, 2, 5, 6, 8} and A= {1, 5} then Ac=U-A= {2, 6, 8}.

4. Difference-
Given two sets A and B, difference of A and B is denoted by A-B.
Given A= {1, 2, 3} and B= {2, 3, 5}
Then A - B= {1}

Properties of Crisp Sets

The following properties are holds on crisp sets.

B=B A and A∩B = B∩A

1. Commutative Law

Associative Law
2. (AUB)UC=AU(BUC) and (A∩B) ∩UC=A∩(B∩C)

(AUB) ∩C=(A∩C)U(B∩C) and (A∩B)UC=(AUB) ∩(BUC)

3. Distributive Law

4. Idempotent Law AUA=A and A∩A=A

5. Identity AUΦ =A, A ,A = and A
 6.
Law of Absorption AU(A∩B)=A

7.
Transitivity A B and B C then A C

8.
Involution (Ac)c =A

9.
Law of Contradiction
A∩Ac =Φ

Law of Excluded middle A U Ac =U

10.

(AUB)c = Ac ∩ Bc. and (A∩B)c =Ac U Bc

11. De Morgan’s Law

Fuzzy Set

Fuzzy sets support a flexible sense of membership of elements to a set. It means in fuzzy set
theory, many degree of membership are allowed. Thus, a membership function µÃ (x) is
associated with set à such that the function maps every element of the universe of discourse X
(or reference set) to the interval [0, 1]. The mapping is written as µÃ(x)  [0, 1]

We will study in next lecture in detail.

Lecture # 3 Fuzzy Set Theory and Operations

Fuzzy Set
Fuzzy sets support a flexible sense of membership of elements to a set. It means in fuzzy set
theory, many degree of membership are allowed. Thus, a membership function µÃ (x) is
associated with set à such that the function maps every element of the universe of discourse X
(or reference set) to the interval [0, 1]. The mapping is written as µÃ(x)  [0, 1]

A fuzzy set is defined as follows:

If X is a universe of discourse and x is a particular element of X, then a fuzzy set à defined on X

may be written as collection of ordered pairs

à = {(x, µÃ(x)): x Є Ã}

Where, each (x, µÃ(x)) is called singleton. In crisp set term µÃ(x) is dropped.

An alternative definition which indicates a fuzzy set as a union of all µÃ(x)/x singletons is given

( )
∑ in discrete case

Ã=∫ in continuous case

EXAMPLE

Let X = {g1, g2, g3, g4, g5} be reference set of students. Let à is fuzzy set of “smart” students

Ã= {(g1, 0.3) (g2, 0.4) (g3, 0.8) (g4, 1) (g5, 0.8)}. It means “smartness” of g1 is 30 %.

Membership Function

Membership function values need not be described using discrete values. It can be described
using continuous values. It can also be describe mathematically as
µÃ(x) = , this function can be shown graphically as in figure 3a

µÃ(x) =1

-3 -2 -1 0 1 2 3
-x x

Figure 3 a Continuous membership function dictated by mathematical function

Different shape of membership function exists; it could be triangular, trapezoidal, curved etc, as
shown figures figure 3 b

Figure 3 b Different shapes of membership function graphs

Solved Example on fuzzy sets

Given set of people in the age group of

0-10 10-20 20-30 30-40

40-50 50-60 60-70 70 and above,
Represent “Young”, “Old” and “Middle-Aged” Person.

Solution. If we plot a graph as in figure 3 d

 Young middle-age old
1

0 10 20 30 40 50 60 70 80 90

Figure 3 d Example of fuzz sets e pressi g ou g , iddle-age , a d ‘age . Age is o horizo tal li e
and membership value on vertical line.

Basic fuzzy set operations

Let X is universe of discourse, Ã and ̃ ar e two fuzzy sets. Let (x) and ̃ (x) are there
respective membership functions.
Basic fuzzy set operations are as follows.

1. Union
Union of two fuzzy sets ̃ and ̃ is new fuzzy set ̃ ̃ with membership

̃ = max ( (x), ̃ (x))

Example

Let à be the fuzzy set of young people and ̃ be the set of middle-aged people. Then à ̃ ,
the fuzzy set of “young or middle-aged “will be given by as

Figure 3 e Union of young and middle-aged person

 In the discrete form, for x1, x2, x3

If ̃ = {(x1, 0.5), (x2, 0.7) (x3, 0)} and ̃ = {(x1, 0.8), (x2, 0.2) (x3, 1)}
̃ ̃ = {(x1, 0.8) (x2, 0.7), (x3, 1)}
Since

̃ (x1)= max ( (x1), ̃ (x1)) =max (0.5, 0.8) = 0.8

̃ (x2)= max ( (x2), ̃ (x2)) =max (0.2.0.7) = 0.7

̃ (x3)= max ( (x3), ̃ (x3)) = max (0, 1) =1

2. Intersection

Intersection of two fuzzy sets ̃ and ̃ is new fuzzy set ̃ ̃ with membership

̃ = min ( (x), ̃ (x))

Example

Let à be the fuzzy set of young people and ̃ be the set of middle-aged people. Then à ̃ ,
the fuzzy set of “young and middle-aged “will be given as in figure 3f

Figure 3 f Intersection of young and middle-aged

 In the discrete form, for x1, x2, x3

If ̃ = {(x1, 0.5), (x2, 0.7) (x3, 0)} and ̃ = {(x1, 0.8), (x2, 0.2) (x3, 1)}
̃ ̃ = {(x1, 0.5) (x2, 0.2), (x3, 0)}
Since

̃ (x1)= min ( (x1), ̃ (x1)) =min (0.5, 0.8) =0.5

̃ (x2)= min ( (x2), ̃ (x2)) =min (0.2.0.7) =0.2

̃ (x3)= min ( (x3), ̃ (x3)) = min (0, 1) =0

3. Complement

The complement of a fuzzy set à is a new fuzzy set ̃ with membership function
(x)=1- (x)
Example. For the fuzzy set A defined as “young” the complement “not young” is given by
as in figure 6 g.

Figure3 g. Complement of “young” is “not young”

In the discrete form, for x1, x2, and x3
If Ã= {(x1, 0.5) (x2, 0.7) (x3, 0)}

Then = {(x1, 0.5) (x2, 0.3) (x3, 1)}

Because

(x1) =1- (x1) =1-0.5 =0.5

(x2) =1- (x2) =1-0.7 =0.7
(x3) =1- (x3) =1-0 =0
Some other operations on fuzzy sets are

4. Product of two fuzzy sets

The product of two fuzzy sets ̃ and ̃ is a new fuzzy set ̃ with membership function

̃ (x)= (x). ̃ (x)

à = {(x1, 0.2), (x2, 0.8) (x3, 0.4)} and ̃ = {(x1, 0.4), (x2, 0) (x3, 0.1)}
̃ = {(x1, 0.08), (x2, 0.)(X3, 0.04)}
Because

̃ (x1)= (x1). ̃ (x1) =0.2.00.4 =0.08

̃ (x2) = (x2). ̃ (x2 ) =0.8.0 =0

̃ (x3) = (x3). ̃ (x3) =0.4.0.1 =0.04

Equality of Two fuzzy sets

Two fuzzy are said to be equal ( ̃ ) if (x) = ̃ (x)

Example Let there are three fuzzy sets ̃ and ̃ as below

Ã= {(x1, 0.2) (x2, 0.8)}

̃ = {(x1, 0.6) (x2, 0.8)}
̃ = {(x1, 0.2) (x2, 0.8)}
 ̃
Since (x1) ≠ ̃ (x1) although (x2) = ̃ (x2)
But ̃
Since (x1) = ̃ (x1) and (x2) = ̃ (x2)

Product of fuzzy set with a crisp number

Multiplying a fuzzy set à by a crisp number b results in a new fuzzy set b.à with the membership
function

(x)=b. (x)

Example

Ã= {(x1, 0.4) (x2, 0.6) (x3, 0.8)} and b=0.3 then

b.Ã = {(x1, 0.12) (x2, 0.18) (x3, 0.24)} since

(x1)=b. (x1)=0.3*0.4=0.12
(x2)=b. (x2)=0.3*0.6=0.18
(x3)=b. (x3)=0.3*0.8=0.24

Power of a fuzzy set

The α power of a fuzzy set à is a new fuzzy set whose membership function is given by
α
(x)=
Raising a fuzzy set to its second power is called Concentration (CON) and taking the square root
is called Dilation (DIL)

Example

Ã= {(x1, 0.16), (x2, 0.2) (x3, 0.7)}

For α=2
(x1)=( )2= (0.16)2=0.0256
 (x2)=( )2= (0.2)2=0.04
2
(x3)= = (0.7)2=0.49

Hence
= {(x1, 0.0256), (x2, 0.04), (x3, 0.49)}

Difference
The difference of two fuzzy sets à and ̃ is new fuzzy set à - ̃ defined as

à -̃= à ̃

Example

Ã= {(x1, 0.2), (x2, 0.5), (x3, 0.6)} and

̃ = {(x1, 0.1), (x2, 0.4), (x3, 0.5)}
̃ = {x1, 0.9), (x2, 0.6), (x3, 0.5)}

à - ̃ =à ̃ = {(x1, 0.2), (x2, 0.5), (x3, 0.5)}

Disjunctive Sum
The disjunctive sum of two fuzzy sets à and ̃ is a new fuzzy set à ̃defined as

à ̃= ̃ ̃) ̃ )
̃ = {(x1, 0.4) (x2, 0.8) (x3, 0.6)}
̃ = {(x1, 0.2) (x2, 0.6) (x3, 0.9)}

= {(x1, 0.6) (x2, 0.2) (x3, 0.4}

̃ = {(x1, 0.8) (x2, 0.4) (x3, 0.1)}
̃ ) = {(x1, 0.4) (x2, 0.4) (x3, 0.1)}
̃ ̃ ) = {(x1, 0.2) (x2, 0.2) (x3, 0.4)}

à ̃= ̃ ̃) ̃ ) = {(x1, 0.4) (x2, 0.4) (x3, 0.4)}

Lecture # 4 Properties of Fuzzy Sets

Fuzzy sets allow some of the properties satisfied by crisp sets. In fact, crisp sets can be thought
of as special instances of fuzzy sets. Any fuzzy set A is subset of the reference set X. Also, the
membership of any element belonging to null set Φ is 0 and the membership of any element
belonging to reference set is 1.

The properties satisfied by fuzzy sets are

Commutative Law : Ã ̃= ̃ Ã and

à ̃= ̃ Ã
Associative Law : ̃ (̃ ̃) = ( ̃ ̃ ) ̃ and
̃ (̃ ̃ ) =( ̃ ̃) ̃

Distributive Law : ̃ (̃ ̃ )= (Ã ̃ ) (Ã ̃) and

̃ (̃ ̃ )=(Ã ̃ ) (Ã ̃)

Law of Idempotent : Ã ̃ = Ã and

à ̃= Ã
Identity : Ã = ,Ã = and
à = ,à =X
Transitivity : à ̃, ̃ ̃ then à ̃

Involution : =Ã
De Morgan, s Law : ( ̃) = ̃ and

( ̃) = ̃

Note-Since fuzzy sets can overlap, the laws of excluded middle and law of
contradiction do not hold good that is à and Ã
Where X is reference set.
Example

The task is to recognize English alphabetical characters (F, E, X, Y, I, T) in an

image processing system.
Define two fuzzy sets ̃ and ̃ to represent the identification of characters I and F.
̃={ }
̃ ={ }
Find the following.

(i) ̃ ̃ (ii) ̃ ̃ (iii) ̃ ̃

Verify De Morgan’s Law ( ̃ ̃ ) =̃ ̃

Solution

(a)
̃ ̃ ={ }

(Membership value of each element in the set ̃ ̃ will be equal to maximum of membership

values of corresponding elements in sets ̃ ̃)

̃ ̃ =̃ ̃ ={ }

Because ̃ ={ }

(Membership value of each element in the set ̃ ̃ will be equal to minimum of membership

values of corresponding elements in sets ̃ ̃ )

̃ ̃ ={ }
(b) De Morgan’s Law (̃ ̃) =̃ ̃

Taking L.H.S. ( ̃ ̃ ) ={ }
={ }

Taking RHS ̃ ̃

̃ ={ }
̃ ={ }
̃ ̃ ={ }

Here LHS=RHS, hence De Morgan’s Law is verified.

Solved Example on fuzzy sets

Consider the fuzzy sets A and b defined on interval X= [0, 5] of real numbers, by the
membership grade functions

(x) = and ̃ (x) =

Determine the mathematical formulae and graphs of the membership grads functions of each of
the following sets
(a) ̌ and ̃ (b) ̃ (c) ̃ (d) ( ̃)

Solution
The membership value for the set ̌ is

(x)=1- (x)

=1- =

And graphically represented as in figure

Figure 4 a Representation of ̌
in figure 4 a
The membership value for the set
̌ is
̃ (x)=1- ̃ (x)

=1-

It can be graphically represented as in

figure4 b
Figure 4 b Representation of ̌

(Note: In figure4a, 4 b, etc graphs are plotted using value x on horizontal line and
corresponding membership values on vertical line)

(b) The membership of ̃ is

determined as

̃ (x) =max ( (x), ̃ (x))

max

It can be graphically represented as

in figure 4 c
Figure 4 c. Representation of ̃

(c) The membership of ̃ is

determined as

̃ (x) =min ( (x), ̃ (x))

min

It can be graphically represented Figure d Representation of ̃

as in figure d
 (d)The membership of ( ̃ ) is is computed using De Morgan’s law i.e.

( ̃) = ̃

=min ( (x), ̃ (x))

=min[ ]

It can be graphically Figure d Representation of ( ̃)

represented as in figure e
 Lecture 5 # Fuzzy and Crisp Relations
CRISP RELATIONS
We will review crisp relations as prelude to fuzzy relations. The concept of relation between sets
is built on the Cartesian product operator of sets.

Cartesian product
The Cartesian product of two sets A and B is denoted by A × B is the set of all ordered pairs
such that the first element in the pair belongs to A and the second element belongs to B.
i.e. A × B = { }
If A B and A and B are non-empty then A × B ≠ B × A

The Cartesian product could be extended to n number of sets

Ai = { }
Observe that
| Ai | = |Ai |

Example
Given
A1= {a, b}, A2= {1, 2}, A3= {α}
A1 × A2={ }, |A1 × A2| = 4 and |A1| = | A2| =2
Here | A1 × A2| = |A1|. | A2|
A1 × A2 × A3={ }

i.e. | A1 × A2 × A3| = | A1|.| A2|.| A3| =4

 Other Crisp Relations
An n-ary relation denoted as R(X1, X2, X3,…, Xn ) among crisp sets X1, X2. ……., Xn is a
subset of artesian product and is indicative of association of relation among the tuple
elements.
For n = 2 the relation R(X, Y) is termed as a binary relation:
For n=3, the relation is termed as ternary;
For n=4, quaternary;
For n=5 quinary and so on.
If the universe of discourse or sets are finite, the n-ary relation can be expressed as an n-
dimensional relation matrix. Thus, for a binary relation R(X, Y) where X = {x1, x2, xn} and Y= {y1,
y2, yn} the relation matrix R is a two dimensional matrix where X represents the rows, Y represents
columns and R (i, j) =1 if ( , ) R and R (i, j) =0 if (xi, yi) R

Example

Given X= {1, 2, 3, 4}

X×X={ }

Let the relation R be defined as R = {(x, y)/ y = x+1, x, y X} on the basis of this formulae then
R is givens as
R={ }

The relation matrix R is given by

[ ]
Operations on Relations

Given two relations and defined on and represented by relation matrices, the following
operations are supported by R and S

Union: :

Intersection: :

Complement: ̅ ̅ (x, y) = 1-R(x, y)

Composition of Relational: R o S

Given R to be a relation on X, Y and S to be a relation on Y, Z then R o S is a composition of

relation on X, Z defined as

RoS={ }
A common form the composition relation is the max-min composition.
Max-min Composition:
Given the relation matrices of the relation R and S, the max-n composition is
defined as
For T=RoS
( )
T(x, y) =

Example
Let R, S be defined on the sets {1, 3, 5} × {1, 3, 5}
Let R: {(x, y) | y = x+2} and S: {(x, y) | x<y}
R = {(1, 3) (3, 5)} and S= {(1, 3) (1, 5) (3, 5)
The relation matrices are
R= S=
[ ] [ ]

Using max-min composition R o S may be computed as

RoS =
[ ]

Because R o S (1, 1) = max (min (0, 0), min (1, 0), min (0, 0))
= max (0, 0, 0)
= 0
In the same way find other values.

Here (R o S (1, 1) is computed using

R o S (1, 1) =max {min (R (1, 1), S (1, 1)), min (R (1, 3), S (3, 1)), min (R (1, 5), S (5, 1))}

Use same method for other computations of Ro S like R o S (1, 3), R o S (1, 5) etc)

Fuzzy Relations

Fuzzy relations is a fuzzy set defined on the Cartesian product of crisp sets X1, X2, …, An
where the n-tuples (x1, x2, …., xn) may have varying degree of membership within the
relation. The membership value indicates the strength of the relation between tuples.

Example

Let R be the fuzzy relation between two sets X1 and X2 where X1 is the set of diseases and X2 is
the set of symptoms.
X1= {typhoid, viral fever, common cold}
X2= {running nose, high temperature, shivering}

The fuzzy relation R may be defined as

Running Nose High Shivering

Temperature
Typhoid 0.1 0.9 0.8
Viral Fever 0.2 0.9 0.7
Common 0.9 0.4 0.6
Cold

(In table degree of membership between relation Typhoid and High Temperature is 0.9
shows strangeness of relation)

Fuzzy Cartesian Product-

Let be fuzzy set defined on the universe and ̃ be a fuzzy set defined on the
universe , the Cartesian product between two fuzzy sets ̃ and ̃ indicated as ̃ ̃and
resulting in a fuzzy relation ̃ is given by
̃= ̃ ̃ X Y
Where ̃ has its membership function given by has its membership function given by
̃ (x, y) = ̃ ̃ (x, y)

= min ( ̃ (x), ̃ (y))

Example
Let à = {(x1, 0.2), (x2, 0.7), (x3, 0.4)} and ̃= {(y1, 0.5), (y2, 0.6)} be two fuzzy sets defined
on the universe of discourse X = [x1, x2, x3} and Y= {y1, y2} respectively. Then the fuzzy
relation R resulting out of the fuzzy Cartesian ̃ ̃ product is given by
̃= ̃ ̃=
[ ]

Since
̃( , y1) = min ( ̃ (x1), ̃ (y1)) =min (0.2, 0.5) = 0.2
̃ (x1, y2) = min ( ̃ (x1), ̃ (y2)) =min (0.2, 0.6) = 0.2
̃ (x2, y1) = min ( ̃ (x2), ̃ (y1)) =min (0.7, 0.5) = 0.5
̃ (x2, y2) = min ( ̃ (x2), ̃( y2)) =min (0.7, 0.6) = 0.6
̃ (x3, y1) = min ( ̃ (x3), ̃( y1)) =min (0.4, 0.5) = 0.4
̃ (X3, y2) = min ( ̃ (x3), ̃ (y2)) =min (0.4, 0.6) = 0.4
Operations on Fuzzy Relations
Let ̃ and ̃ be fuzzy relations on X ×Y
Union : ̃ ̃

̃ ̃ (x, y) = max ( ̃ (x, y), ̃ (x, y))

Intersection: ̃ ̃

̃ ̃ (x, y) = min ( ̃ (x, y), ̃ (x, y))

Complement: ̃
̃ (x, y) = 1- ( ̃ (x, y)

Composition of Relations
The definition is similar to crisp relation. Suppose ̃ is a fuzzy relation defined on X Y, and ̃
is a fuzzy relation defined on Y Z then ̃ o ̃ is a fuzzy relation on X ×Z. The fuzzy max-min
composition is defined as

̃ ̃ (x, z) = ̃ ̃
Example
X={ } Y={ } Z={ }

Let ̃ be a fuzzy relation on X ,̃

[ ]

Let ̃ be a fuzzy relation Y ,̃ =

[ ]

Then ̃ o ̃ on X , by max-min composition yields,

[ ]

Because
̃ ̃ ( , )= ̃ ̃ ̃ ̃

= max (min (0.5, 0.6), min (0.1, 0.5))

= max (0.5, 0.1) = 0.5
̃ ̃ ( , )= ( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.5, 0.4), min (0.1, 0.8))
= max (0.4, 0.1) = 0.4

̃ ̃ ( , )= ( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.5, 0.7), min (0.1, 0.9))
= max (0.5, 0.1) = 0.5

̃ ̃ ( , )= ( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.2, 0.6), min (0.9, 0.5))
= max (0.2, 0.5) = 0.5
 ̃ ̃ ( , ) =

( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.2, 0.4), min (0.9, 0.8))
= max (0.2, 0.8) = 0.8
̃ ̃ ( , ) =
( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.2, 0.7), min (0.9, 0.9))
= max (0.2, 0.9) = 0.9

̃ ̃ ( , ) =
( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.8, 0.6), min (0.6, 0.5))
= max (0.6, 0.5) = 0.5

̃ ̃ ( , ) = ( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.8, 0.4), min (0.6, 0.8))
= max (0.4, 0.6) = 0.6

̃ ̃ ( , ) = ( ( ̃ ) ̃ ( ̃ ) ̃ )
= max (min (0.8, 0.7), min (0.6, 0.9))
= max (0.7, 0.6) = 0.

Numerical on Fuzzy max-min Composition

Consider a set P ={ } of four varieties of paddy plants, set D={ } of

various diseases affecting the plants and S={ } be the common symtoms of the
disease.
Let ̃ be a relation on P × D and ̃ be relation on D × S For

̃= and ̃=
[ ] [ ]

Obtain the association of the plants with the different symptoms of the diseases
using max-min composition.

Solution
To obtain the association of the plants with the symptoms, ̃ o ̃ which is a relation on the set?
P S is computed as

Using max-min composition,

̃õ=
[ ]

Because
̃ ̃( = max (min ( ̃ ( , ̃( , min ( ̃ ( , ̃( ,
min ( ̃ ( , ̃( , min ( ̃ ( , ̃(

=max (min (0.6, 0.1), min (0. 6, 1), min (0.9, 0), min (0.8, 0.9))
= max (0.1, 0.6, 0, 0.8) = 0.8

̃ ̃( = max (min ( ̃ ( , ̃( , min ( ̃ ( , ̃( ,

min ( ̃ ( , ̃( , min ( ̃ ( , ̃(
 =max (min (0.6, 0.2), min (0. 6, 1), min (0.9, 0), min (0.8, 1))
= max (0.2, 0.6, 0, 0.8) = 0.8

̃ ̃( = max (min ( ̃ ( , ̃( , min ( ̃ ( , ̃( ,

min ( ̃ ( , ̃( , min ( ̃ ( , ̃(

=max (min (0.6, 0.7), min (0. 6, 4), min (0.9, 0.5), min (0.8, 0.8))
= max (0.6, 0.4, 0.5, 0.8) = 0.8
Similarly ̃ ̃( =0.9, ̃ ̃( =0.8 ̃ ̃( =0.8
̃ ̃( =0.8, 8 ̃ ̃( =0.9 and so on as in matrix above