Defuzzification Techniques

Soft Computing Applications

What is defuzzification?

Defuzzification means the fuzzy to crisp conversion.

Example 1:
Suppose, THIGH denotes a fuzzy set representing temperature is
High.
THIGH is given as follows.
THIGH = (15,0.1), (20, 0.4), (25,0.45), (30,0.55), (35,0.65),
(40,0.7), (45,0.85),(50,0.9)

What is the crisp value that implies for the high temperature?

Soft Computing Applications

Example 2: Fuzzy to crisp

As an another example, let us consider a fuzzy set whose membership

finction is shown in the following figure.

 (x)

What is the crisp value of the fuzzy set in this case?

Soft Computing Applications

Example 3: Fuzzy to crisp
Now, consider the following two rules in the fuzzy rule base.

R1: If x is A then y is C
R2: If x is B then y is D
A pictorial representation of the above rule base is shown in the
following figures.

C
1.0 1.0
A B

 x’
 D

x y

What is the crisp value that can be inferred from the above rules given
′
an input say x ?
Soft Computing Applications
Why defuzzification?

The fuzzy results generated can not be used in an application, where

decision has to be taken only on crisp values.

Example:

If THIGH then rotate RFIRST .

Here, may be input THIGH is fuzzy, but action rotate should be based
on the crisp value of RFIRST .

Soft Computing Applications

Generic structure of a Fuzzy system

Following figures shows a general fraework of a fuzzy system.

Fuzzy
rule
Crisp Defuzzifier Crisp
Fuzzifier base
input output

Inference
mechanism

Soft Computing Applications

Defuzzification Techniques

Soft Computing Applications

Defuzzification methods

A number of defuzzification methods are known. Such as

1 Lambda-cut method

2 Weighted average method

3 Maxima methods

4 Centroid methods

Soft Computing Applications

Lambda-cut method

Soft Computing Applications

Lambda-cut method

Lmabda-cut method is applicable to derive crisp value of a fuzzy set or

relation. Thus

Lambda-cut method for fuzzy set

Lambda-cut method for fuzzy relation

In many literature, Lambda-cut method is also alternatively termed as

Alph-cut method.

Soft Computing Applications

Lamda-cut method for fuzzy set

1 In this method a fuzzy set A is transformed into a crisp set Aλ for a

given value of λ (0 ≤ λ ≤ 1)

2 In other-words, Aλ = { x|µA (x) ≥ λ}

3 That is, the value of Lambda-cut set Aλ is x, when the

membership value corresponding to x is greater than or equal to
the specified λ.

4 This Lambda-cut set Aλ is also called alpha-cut set.

Soft Computing Applications

Lambda-cut for a fuzzy set : Example

A1 = { (x1,0.9),(x2,0.5),(x3,0.2),(x4,0.3)}

Then A0.6 = {(x 1, 1), (x2, 0), (x3, 0), (x4, 0)} = {x 1 }

and

A2 = { (x1,0.1),(x2,0.5),(x3,0.8),(x4,0.7)}

A0.2 = { (x1 ,0),(x2,1),(x3 ,1),(x4,1)} = { x2,x3,x4}

Soft Computing Applications

Lambda-cut sets : Example
Two fuzzy sets P and Q are defined on x as follows.

µ(x) x1 x2 x3 x4 x5
P 0.1 0.2 0.7 0.5 0.4
Q 0.9 0.6 0.3 0.2 0.8

Find the following :

(a) P0.2, Q0.3

(b) (P ∪Q) 0.6

(c) (P ∪P) 0.8

(d) (P ∩Q) 0.4

Soft Computing Applications

Lambda-cut for a fuzzy relation

The Lambda-cut method for a fuzzy set can also be extended to fuzzy
relation also.
Example: For a fuzzy relation R
1 0.2 0.3
R = 0.5 0.9 0.6
0.4 0.8 0.7
We are to find λ-cut relations for the following values of
λ = 0,0.2,0.9,0.5
1 1 1 1 1 1
R0 = 1 1 1 and R0.2 = 1 1 1 and
1 1 1 1 1 1
1 0 0 1 0 0
R0.9 = 0 1 0 and R0.5 = 1 1 1
0 0 0 0 1 1

Soft Computing Applications

Some properties of λ-cut sets

If A and B are two fuzzy sets, defined with the same universe of
discourse, then

1 (A ∪B) λ = Aλ ∪Bλ

2 (A ∩B) λ = Aλ ∩Bλ

3 (A) λ /= Aλ except for value of λ = 0.5

4 For any λ ≤ α, where α varies between 0 and 1, it is true that

Aα ⊆ Aλ , where the value of A0 will be the universe of discourse.

Soft Computing Applications

Some properties of λ-cut relations

If R and S are two fuzzy relations, defined with the same fuzzy sets
over the same universe of discourses, then

5 (R ∪S) λ = Rλ ∪Sλ

6 (R ∩S) λ = Rλ ∩Sλ

7 (R) λ /= R λ

8 For λ ≤ α, where α between 0 and 1 , then Rα ⊆ Rλ

Soft Computing Applications

Summary: Lambda-cut methods

Lambda-cut method converts a fuzzy set (or a fuzzy relation) into crisp
set (or relation).

Soft Computing Applications

Output of a Fuzzy System

Soft Computing Applications

Output of a fuzzy System

The output of a fuzzy system can be a single fuzzy set or union of two
or more fuzzy sets.

To understand the second concept, let us consider a fuzzy system with

n-rules.
R1: If x is A1 then y is B1
R2: If x is A2 then y is B2
........................................
........................................
Rn: If x is An then y is Bn
In this case, the output y for a given input x = x1 is possibly B =
B1 ∪B2 ∪.....Bn

Soft Computing Applications

Output fuzzy set : Illustration
Suppose, two rules R1 and R2 are given as follows:

1 R1: If x is A1 then y is C1
2 R2: If x is A2 then y is C2

Here, the output fuzzy set C = C1 ∪C2.

For instance, let us consider the following:

C2
1.0 1.0
A C1


B


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

x y
x1 x2 x3

Soft Computing Applications

Output fuzzy set : Illustration

The fuzzy output for x = x1 is shown below.

1.0 1.0
A C
B 


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

x1 y
x
Fuzzy output for x = x1

Soft Computing Applications

Output fuzzy set : Illustration

The fuzzy output for x = x2 is shown below.

1.0 1.0
A

B  C


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

x = x2 y
x
Fuzzy output for x = x2

Soft Computing Applications

Output fuzzy set : Illustration

The fuzzy output for x = x3 is shown below.

1.0 1.0
A
B  C


0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 8
y
x = x3
x
Fuzzy output for x = x3

Soft Computing Applications

Defuzzification Methods
Following defuzzification methods are known to calculate crisp output
in the situations as discussed in the last few slides
Maxima Methods
1 Height method
2 First of maxima (FoM)
3 Last of maxima (LoM)
4 Mean of maxima(MoM)

Centroid methods
1 Center of gravity method (CoG)
2 Center of sum method (CoS)
3 Center of area method (CoA)

Weighted average method

Soft Computing Applications

Defuzzification Technique
Maxima Methods

Soft Computing Applications

Maxima methods
Following defuzzification methods are known to calculate crisp output.
Maxima Methods
1 Height method
2 First of maxima (FoM)
3 Last of maxima (LoM)
4 Mean of maxima(MoM)

Centroid methods
1 Center of gravity method (CoG)
2 Center of sum method (CoS)
3 Center of area method (CoA)

Weighted average method

Soft Computing Applications

Maxima method : Height method
This method is based on Max-membership principle, and defined as
follows.

µC (x ∗ ) ≥ µC (x) for all x ∈ X

c

Note:
1. Here, x ∗ is the height of the output fuzzy set C.
2. This method is applicable when height is unique.
Soft Computing Applications
Maxima method : FoM

FoM: First of Maxima : x ∗ = min{ x|C(x) = maxw C{ w} }

c

Soft Computing Applications

Maxima method : LoM

LoM : Last of Maxima : x ∗ = max{ x|C(x) = maxw C{ w} }

c

Soft Computing Applications

Maxima method : MoM

Σ
(xi )
x∗ = xi ∈M
|M|

where, M = {x i |µ(xi ) = h(C)} where h(C) is the height of the fuzzy set
C

Soft Computing Applications

MoM : Example 1

Suppose, a fuzzy set Young is defined as follows:

Young = {(15,0.5), (20,0.8), (25,0.8), (30,0.5), (35,0.3) }

Then the crisp value of Young using MoM method is

x∗ = 20+25
2
= 22.5

Thus, a person of 22.5 years old is treated as young!

Soft Computing Applications

MoM : Example 2
What is the crisp value of the fuzzy set using MoM in the following
case?

c

x∗ = a+b
2

Note:
Thus, MoM is also synonymous to middle of maxima.
MoM is also general method of Height.
Soft Computing Applications
Defuzzification Technique
Centroid Methods

Soft Computing Applications

Cenroid methods
Following defuzzification methods are known to calculate crisp output.
Maxima Methods
1 Height method
2 First of maxima (FoM)
3 Last of maxima (LoM)
4 Mean of maxima(MoM)

Centroid methods
1 Center of gravity method (CoG)
2 Center of sum method (CoS)
3 Center of area method (CoA)

Weighted average method

Soft Computing Applications

Centroid method : CoG
1 The basic principle in CoG method is to find the point x ∗ where a
vertical line would slice the aggregate into two equal masses.
2 Mathematically, the CoG can be expressed as follows :
,
x.µC (x)dx
x∗ = ,
µC (x)dx

3 Graphically,

Center of gravity

c

x*
x

Soft Computing Applications

Centroid method : CoG

Note:
1 x ∗ is the x-coordinate of center of gravity.
∫
2 µC (x)dx denotes the area of the region bounded by the curve
µC .

3 If µC is defined with a discrete membership function, then CoG

can be stated as :
Σn
xi .µC (xi )
x∗ = Σi=1 ;
i=1 µC (xi )
n

4 Here, xi is a sample element and n represents the number of

samples in fuzzy set C.

Soft Computing Applications

CoG : A geometrical method of calculation
Steps:
1 Divide the entire region into a number of small regular regions
(e.g. triangles, trapizoid etc.)

A4
A3
A2
A1 A5
A6

x1 x2 x3 x4 x5 x6
x

2 Let Ai and xi denotes the area and c.g. of the i-th portion.
3 Then x ∗ according to CoG is
Σn
x∗ = i=1 xi .(Ai )
Σ n
i=1 Ai
where n is the number of smaller geometrical components.
Soft Computing Applications
CoG: An example of integral method of calculation

c2
1.0 1.0
c1
0.7 0.7
c
 c 0.5
1
2

0.5

0 1 2 3 4 5 0 1 2 3 4 5 6
x x
C = C1  C 2
d e
1.0
b c
0.7
c
0.5
A1 A2 A3 A4 A5

f
a
0 1 2 2.7 3 4 5 6

Soft Computing Applications

CoG: An example of integral method of calculation

0.35x 0≤x <2

0.7 2 ≤ x < 2.7
µc (x) = x −2 2.7 ≤ x < 3
1 3≤x <4
(−0.5x + 3) 4≤x ≤6

For A1 : y − 0 = 0.7 (x
2
− 0), or y = 0.35x

For A2 : y = 0.7
For A3 : y − 0 = 1− 0 (x
3− 2
− 2), or y = x − 2
For, A4 : y = 1
For, A5 : y − 1 = 0− 1 (x
6− 4
− 4), or y = −0.5x + 3

Soft Computing Applications

CoG: An example of integral method of calculation

,
x.µ (x)dx
Thus, x ∗ = , c = N
µc (x)dx D

∫ ∫ ∫3 ∫
N = 02 0.35x 2dx + 22.7 0.7x 2dx + 2.7 (x 2 − 2x)dx + 34 xdx +
∫6
4
(−0.5x 2 + 3x)dx

= 10.98
∫2 ∫ 2.7 ∫3 ∫4 ∫6
D= 0
0.35xdx + 2
0.7xdx + 2.7
(x −2)dx + 3
dx + 4
(−0.5x +3)dx

= 3.445

Thus, x ∗ = 10.98
3.445
= 3.187

Soft Computing Applications

Centroid method : CoS
If the output fuzzy set C = C1 ∪C2 ∪....Cn , then the crisp value
according to CoS is defined as

Σn
x∗ = i=1 xi .Aci
Σn
i=1 Aci

Here, Aci denotes the area of the region bounded by the fuzzy set Ci
and xi is the geometric center of the area Aci .

Graphically,

c2
c3
c1
A2 A3
A1
5

x1 x2 x3

Soft Computing Applications

Centroid method : CoS

Note:

1 In CoG method, the overlapping area is counted once, whereas, in

CoS , the overlapping is counted twice or so.

2 In CoS, we use the center of area and hence, its name instead of
center of gravity as in CoG.

Soft Computing Applications

CoS: Example

Consider the three output fuzzy sets as shown in the following plots:

c 1 1.0
c 2
0.5 0.5
 c3 0.5
0.3
0.25 0.25 0.25

0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
x x x

Soft Computing Applications

CoS: Example

c 1.0
c
1

0.5
2

0.5
c 3 0.5
0.3
0.25 0.25 0.25

0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
x x x

In this case, we have

Ac1 = 1
2
× 0.3 × (3 + 5),x1 = 2.5
Ac2 = 1
2
× 0.5 × (4 + 2),x2 = 5
Ac3 = 1
2
× 1 × (3 + 1),x3 = 6.5
1
×0.3×(3+5)×2.5+ 12× 0 .5×(4+2)×5+ 12×1×(3 +1)×6.5
Thus, x ∗ = 2
1 = 5.00
2
×0.3×(3+5+ 12×0.5 ×(4+2)+ 12× 1 × (3+1)

Note:
The crisp value of C = C1 ∪C2 ∪C3 using CoG method can be found
to be calculated as x ∗ = 4.9
Soft Computing Applications
Centroid method: Certer of largest area
If the fuzzy set has two subregions, then the center of gravity of the
subregion with the largest area can be used to calculate the
defuzzified value.
, ′
µ (x).x dx
Mathematically, x ∗ = , cm ;
µcm (x)dx
′
Here, Cm is the region with largest area, x is the center of gravity of
C m.
Graphically,

C1 C3
C2

Cm = C3 x'

Soft Computing Applications

Weighted Average Method

Soft Computing Applications

Cenroid methods
Following defuzzification methods are known to calculate crisp output.
Maxima Methods
1 Height method
2 First of maxima (FoM)
3 Last of maxima (LoM)
4 Mean of maxima(MoM)

Centroid methods
1 Center of gravity method (CoG)
2 Center of sum method (CoS)
3 Center of area method (CoA)

Weighted average method

Soft Computing Applications

Weighted average method

1 This method is also alternatively called ”Sugeno defuzzification”

method.
2 The method can be used only for symmetrical output membership
functions.
3 The crisp value accroding to this method is
Σn
i=1 µC i (xi ).(xi )
x =
∗
Σ
i=1 µC i (xi )
n

where, C1, C2, ...Cn are the output fuzzy sets and (xi ) is the value
where middle of the fuzzy set Ci is observed.

Soft Computing Applications

Weighted average method

Graphically,

k1 C3
k2 C1
C2
k3

x1 x2 x3

Soft Computing Applications

Exercise 1

Find the crisp value of the following using all defuzzified methods.

C2
1.0

C1
0.5

1 2 3 4 5 6

Soft Computing Applications

Exercise 1

Find the crisp value of the following using all defuzzified methods.

C2
1.0

C3
0.75

C1
0.5

1 2 3 4 5 6 7 8 9 10

Soft Computing Applications

Exercise 3

The membership function defining a student as Average, Good,

and Excellent denoted by respective membership functions are as
shown below.

Avg Good Excellent

1.0

0.5

6.0 6.5 7 7.5 8.0 8.5 9.0 10.0

Find the crisp value of ”Good Student”

Hint: Use CoG method to the portion ”Good” to calculate it.

Soft Computing Applications

Exercise 4

narrow wide
1.0

0.5

0.4

5 6 7 8 9 10

The width of a road as narrow and wide is defined by two fuzzy

sets, whose membership functions are plotted as shown above.

If a road with its degree of membership value is 0.4 then what will
be its width (in crisp) measure.

Hint: Use CoG method for the shadded region.

Soft Computing Applications

Exercise 5
The faulty measure of a circuit is defined fuzzily by three fuzzy
sets namely Faulty(F), Fault tolerant (FT) and Robust(R) defined
by three membership functions with number of faults occur as
universe of discourses and is shown below.
 (x)  (x)  (x)

1.0 1.0
1.0 1.0

0.75 0.75 0.75

0.5 0.5
0.5 0.5
0.3

0.25 0.25 0.25

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

Robust Fault tolerant Faulty

Reliability is measured as R∗ = F ∪FT ∪R.

With a certain observation in testing
(x,0.3) ∈ R,(x,0.5) ∈ FT ,(x,0.8) ∈ F .
Calculate the reliability measure in crisp value.
Calculate with 1) CoS 2) CoG .
Soft Computing Applications
Any questions??

Soft Computing Applications