Chapter 3

Fuzzy Membership Function Formulation and Parameterization

The membership function of a fuzzy set is a generalization of the indicator function in classical
sets. In fuzzy logic, it represents the degree of truth as an extension of valuation. Degrees of truth
are often confused with probabilities, although they are conceptually distinct, because fuzzy truth
represents membership in vaguely defined sets, not likelihood of some event or condition.
Membership functions were introduced by Zadeh in the first paper on fuzzy sets (1965).
 

Figure 3.1: Membership functions in Fuzzy vs. crisp sets

Formal Definition of membership function

Let us consider fuzzy set A, A = {(x, µA(x))| x Є X} where µA(x) is called the membership
function for the fuzzy set A. X is referred to as the universe of discourse. The membership
function associates each element x Є X with a value in the interval [0, 1].

In fuzzy sets, each elements is mapped to [0,1] by membership function. That is, μA : X €[0, 1],
where [0,1] means real numbers between 0 and 1 (including 0,1). Consequently, fuzzy set is with
‘vague boundary set’ comparing with crisp set.

The fuzzy set A can be alternatively denoted as follows:

If X is discrete then A = Σ μA(xi) / xi

1
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
If X is continuous then A = ʃ μA(x) / x

Here, μA (x) is the “membership function”. Value of this function is between 0 and 1. This value
represents the “degree of membership” (membership value) of element x in set A. The members
of a fuzzy set are members to some degree, known as a membership grade or degree of
membership. The membership grade is the degree of belonging to the fuzzy set. The larger the
number (in [0, 1]) the more the degree of belonging. (N.B. This is not a probability).The
translation from x to µA(x) is known as fuzzification.

In the fuzzy theory, fuzzy set A of universe X is defined by function µA(x) called the
membership function of set A. We already discussed this point.

µA(x): X → [0, 1], where µA(x) = 1 if x is totally in A;

µA(x) = 0 if x is not in A;

0 < µA(x) < 1 if x is partly in A.

This set allows a continuum of possible choices. For any element x of universe X, membership
function µA(x) equals the degree to which x is an element of set A. This degree, a value between
0 and 1, represents the degree of membership, also called membership value, of element x in set
A.

Basics on Fuzzy Membership Functions

Support: The support of fuzzy set A is the set of all point x  X such that A(x) > 0
Mathematically we can express Support (A) = {(x,A(x))/ A(x)>0)}

Core: The core of a fuzzy set A is the set of all xX such that A(x) = 1.
Mathematically we can express core (A) = {x  X | A(x) = 1}

Crossover: A crossover point of a fuzzy set ‘A’ is a point x  X at which A(x) = 0.5
Mathematically we can express Crossover (A) = {x  X | A(x) = 0.5}

Normality: A fuzzy set ‘A’ is a normal if its core is non-empty, i.e. core (A)    A is a normal fuzzy
set.
Mathematically we can express Normality (A) =1 if A(x) =1, for all xX and (x, A(x))  A

2
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
Fuzzy singleton: A fuzzy set ‘A’ whose support is single point in x with A(x) =1 is called fuzzy
singleton.
|A|=| {(x, A(x))| A(x) =1}|

 - cut: A = {x  X | A(x)  }

Strong  - cut: A = {x  X | A(x) >}. In this case ‘A’ is defined as Crisp set.

Convexity of Fuzzy Sets: A fuzzy set A is convex if and only if for any x1, x2 X and there exists x =
[0, 1] such that

 A (x1  (1   ) x2 )  min(  A ( x1 ),  A ( x2 ))
Alternatively, A is convex if all its -cuts are convex.

Bandwidths: for a normal & convex set, the bandwidth is the distance between two unique crossover
points Bandwidth (A) = |x2 – x1| where A(x1) = A(x2) = 0.5.

Symmetry: A fuzzy set A is symmetric if it’s MF around a certain point x = c, satisfies the following
criteria i.e. A(x + c) = A(c – x) x  X.

Open left, Open Right and Closed:

A fuzzy set ‘A’ is open left A  lim  A ( x )  1 and lim  A ( x)  0
x -  x  

A fuzzy set ‘A’ is open right A  lim  A ( x)  0 and lim  A ( x)  1

x -  x  

A fuzzy set ‘A’ is closed A  lim  A ( x)  lim  A ( x)  0

x -  x  
 

Figure 3.2: Illustration of terminologies on Fuzzy Membership function

Membership functions: Parameterization and Formulation

1) Triangular Membership function

3
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
2) Trapezoidal MF
3) Gaussian MF
4) Generalized bell MF
5) Sigmoid membership function

 
Figure 3.3: Various type of Fuzzy membership functions

Triangular Membership function:

Let a, b and c represent the x coordinates of the three vertices of μ A (x) in a fuzzy set A (a: lower
boundary and c: upper boundary where membership degree is zero, b: the centre where membership
degree is 1).

   
      1.0  0 if x  a 
x  a 
   if a  x  b
 
 A ( x)   b  a
 

μ  (x) 
A c  x if b  x  c 
  c  b 
  0 if x  c 
a          b                   c 
 

4
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
Trapezoidal membership function:

Let a, b, c and d represents the x coordinates of the membership function. then

Trapezoid(x; a, b, c, d) = 0 if x  a;

= (x-a)/ (b-a) if a  x  b

= 1 if b  x  c; 1.0

= (d-x)/ (d-c) 0 if c  x  d;  A (x)

= 0, if d  x.

a b c d

  xa d x 
μ trapezoid = max min  ,1, ,0 
 ba d c  

Gaussian membership function:

The Gaussian membership function is usually represented as Gaussian(x:c,s) where c, s represents the
mean and standard deviation.
 1 xc m
 A ( x, c, s, m)  exp  
 2 s 
Here c represents centre, s represents width and m represents fuzzification factor.

0.9
1
0.8 1
0.9

0.7 0.9
0.8
0.8
0.6 0.7
0.7
0.5 0.6
0.6
0.5
0.4 0.5
0.4
0.3 0.4
0.3
0.3
0.2
0.2
0.2
0.1
0.1 0.1

0 0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10

 
c=5, s=0.5, m=2  c=5, s=2, m=2  c=5, s=5, m=2 
 

Figure 3.4: Different shapes of Gaussian MFs with different values of s and m.

5
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
Generalized Bell membership function:
A generalized bell membership function has three parameters: a –responsible for its width, c –
responsible for its center and b –responsible for its slopes. Mathematically,

1
gbellmf ( x; a, b, c)  2b
xc
1
b

Figure 3.5: Different shapes of Gaussian MFs with different values of s and m.

Sigmoid Membership function:

A sigmoidal membership function has two parameters: a responsible for its slope at the crossover
point x = c. The membership function of the sigmoid function can be represented as Sigmf (x:a, c) and it
is 1
sigmf ( x; a, b, c) 
1  e  a ( x c )

6
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
 Figure 3.6: A general structures of sigmoid MF.

A sigmoidal MF is inherently open right or left & thus, it is appropriate for representing concepts
such as “very large” or “very negative”. Sigmoidal MF mostly used as activation function of
artificial neural networks (NN).A NN should synthesize a close MF in order to simulate the
behavior of a fuzzy inference system.

Left –Right (LR) MF

 cx
 FL   , x  c

LR ( x ; c ,  ,  )    
 F R  x  c  , x  c
   
Example:
3
F L ( x )  m ax ( 0 ,1  x 2
) ,F R ( x )  exp( x )

c=65 c=25
a=60 a=10

Figure 3.7: Examples of L-R MFs

 

 
7
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
2-D membership function
In this case, there are two inputs assigned to an MF: this MF is a two dimensional MF. A one
input MF is called ordinary MF.

Extension of a one-dimensional MF to a two-dimensional MF via cylindrical extensions is

shown in below fig. If A is a fuzzy set in X, then its cylindrical extension in X*Y is a fuzzy set
C(A) defined by: C(A) can be viewed as a two-dimensional fuzzy set.

C ( A)  
X *Y
A ( x ) | ( x, y )

Base set Cylindrical extension

Figure 3.8: Example of a 2D MF

 

Projection of fuzzy sets (decrease dimension): Let R be a two-dimensional fuzzy set on X*Y.
Then the projections of R onto X and Y are defined as:
R X   max  R ( x, y ) | x
 y 

RY   max  R ( x, y ) | y
x

X Y

Figure 3.9: Different views of 2D MFs

8
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
Composite anf non-composite MFs
Suppose that the fuzzy A = “(x,y) is near (3,4)” is defined by:
  x  3 2 
 A ( x, y )  exp      y  4 
2

  2  
  x  3 2    y  4 2 
 exp     exp    
  2     1  
 G ( x;3,2) * G ( y;4,1)

This two-dimensional MF is composite, the fuzzy set A is composed of two statements:

“x is near 3” and “y is near 4”

These two statements are respectively defined as:  near 3 (x) = G(x;3,2) &  near 4 (x) = G(y;4,1)

If a fuzzy set is defined by:

1
 A ( x, y )  ,it is non-composite.
1 | x  3 || y  4 |2.5
A composite two-dimensional MF is usually the result of two statements joined by the AND or
OR connectives.

Composite two-dimensional MFs based on min & max operations

Let trap(x) = trapezoid (x;-6,-2,2,6) ,trap(y) = trapezoid (y;-6,-2,2,6) be two trapezoidal MFs on
X and Y respectively. By applying the min and max operators, we obtain two-dimensional MFs
on X*Y.

Figure 3.9: Two dimensional MFs defined by the min and max operators
9
 
©Debasis Samanta, Indian Institute of Technology Kharagpur 
 
References
 

1. Leszek Rutkowski., Flexible Neuro-Fuzzy Systems: Structures, Learning and

Performance Evaluation, Technical University of Czestochowa, Poland, Kluwer
Academic Publishers New York,   Chapter 2:ELEMENTS OF THE THEORY OF
FUZZY SETS.

2. Jyh-Shing Roger Jang et al., Neuro-Fuzzy and Soft Computing: A Computational

Approach to Learning and Machine Intelligence, First Edition, Prentice Hall, 1997,
Chapter 2: Fuzzy Sets.

3. Bilgic, Taner and Turksen, Burhan. 1999. Measurement of Membership Function:

Theoretical and Experimental Work. In D. Dubois and H. Prade (editors) Handbook of
Fuzzy Systems, Vol. 1, Fundamentals of Fuzzy Sets, chapter 3, pages 195--202. Kluwer
Academic Publishers.

4. W. Barada and H. Singh, Generating Optimal Adaptive Fuzzy-Neural Models of

Dynamical Systems with Applications to Control, IEEE Transactions on Systems, Man,
and Cybernetics, Part C 28 (1998) 371-391.

10
 
©Debasis Samanta, Indian Institute of Technology Kharagpur