Chapter 2

The Operation of Fuzzy Set

2.1 Standard operations of fuzzy set

Complement set A
 A ( x) = 1 −  A ( x)
Union A  B
 AB ( x) = Max[ A ( x), B ( x)]
Intersection A  B
 AB ( x) = Min [ A ( x), B ( x)]
difference between characteristics of crisp fuzzy set operator
◼ law of contradiction A A = 
◼ law of excluded middle A A = X
Table 2.1 Characteristics of standard fuzzy set operators

(1) Involution
A=A
(2) Commutativity AB=BA
A B=B A
(3) Associativity (A  B)  C = A  (B  C)
(A  B)  C = A  (B  C)
(4) Distributivity A  (B  C) = (A  B)  (A  C)
A  (B  C) = (A  B)  (A  C)

(5) Idempotency AA=A

A A = A
(6) Absorption A  (A  B) = A
A  (A  B) = A
(7) Absorption by X and  AX=X
A  =
(8) Identity A=A
A X = A
(9) De Morgan’s law
A B = A  B A B = A  B
(10) Equivalence formula
( A  B)  ( A  B ) = ( A  B )  ( A  B)
(11) Symmetrical difference formula
( A  B)  ( A  B ) = ( A  B )  ( A  B)
2.2 Fuzzy complement

2.2.1 Requirements for complement function

◼ Complement function
C: [0,1] → [0,1]

 A ( x) = C ( A ( x))
(Axiom C1) C(0) = 1, C(1) = 0 (boundary condition)
(Axiom C2) a,b  [0,1]
if a  b, then C(a)  C(b) (monotonic non-increasing)
(Axiom C3) C is a continuous function.
(Axiom C4) C is involutive.
C(C(a)) = a for all a  [0,1]
2.2 Fuzzy complement

2.2.2 Example of complement function(1)

C(a)

C(a) = 1 - a

1 a

Fig 2.1 Standard complement set function

 2.2 Fuzzy complement

2.2.2 Example of complement function(2)

◼ standard complement set function
 A (x)
 A (x)
A A
1 1

1 x 1 x
2.2 Fuzzy complement

2.2.2 Example of complement function(3)

C(a)
1 for a  t
C (a) = 
1 0 for a  t

It does not hold C3 and C4

t 1 a
2.2 Fuzzy complement

2.2.2 Example of complement function(4)

Continuous fuzzy complement function C(a) = 1/2(1+cosa)
1.0
0.9
C(a) 0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
2.2 Fuzzy complement

2.2.2 Example of complement function(5)

◼ Yager complement function
1.0
w=5
0.9
Cw(a) 0.8
w=2
0.7 Cw (a) = (1 − a w )1 / w
0.6
w=1 w  (−1, )
0.5
0.4
0.3 w=0.5
0.2
0.1
0.0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
a
2.2 Fuzzy complement

2.2.3 Fuzzy Partition

( A1 , A2 , , Am )

(1) i, Ai  
(2) Ai  A j = , for i  j
m
(3) x  X , 
i =1
Ai ( x) = 1
2.3 Fuzzy union

2.3.1 Axioms for union function

U : [0,1]  [0,1] → [0,1]
AB(x) = U[A(x), B(x)]

(Axiom U1) U(0,0) = 0, U(0,1) = 1, U(1,0) = 1, U(1,1) = 1

(Axiom U2) U(a,b) = U(b,a) (Commutativity)
(Axiom U3) If a  a’ and b  b’, U(a, b)  U(a’, b’)
Function U is a monotonic function.
(Axiom U4) U(U(a, b), c) = U(a, U(b, c)) (Associativity)
(Axiom U5) Function U is continuous.
(Axiom U6) U(a, a) = a (idempotency)
2.3 Fuzzy union

2.3.2 Examples of union function

U[A(x), B(x)] = Max[A(x), B(x)], or AB(x) = Max[A(x), B(x)]

A B
1 1

X X
AB
1

X
Fig 2.6 Visualization of standard union operation
2.3 Fuzzy union
Yager’s union function :holds all axioms except U6.
U w (a, b) = Min [1, (a w + b w )1/ w ] where w  (0, )
0 0.25 0.5
a
1 1 1 1 U1(a,b) = Min[1, a+b]

0.75 0.75 1 1

0.25 0.25 0.5 0.75 w=1

0 0.25 0.5
a
1 1 1 1 U2(a,b) = Min[1, a 2 + b 2 ]

0.75 0.75 0.79 0.9

0.25 0.25 0.35 0.55 w=2

0 0.25 0.5
a
1 1 1 1 U(a,b) = Max[ a, b] : standard union function

0.75 0.75 0.75 0.75

0.25 0.25 0.25 0.5 w→

 2.3.3 Other union operations

1) Probabilistic sum A+̂ B (Algebraic sum)

x  X ,  A+ˆ B ( x) =  A ( x) + B ( x) −  A ( x)B ( x)

◼ commutativity, associativity, identity and De Morgan’s law

◼
A +̂ X = X

2) Bounded sum AB (Bold union)

x  X ,  A B ( x) = Min [1,  A ( x) + B ( x)]
◼ Commutativity, associativity, identity, and De Morgan’s Law
◼ A X = X , A A = X
◼ not idempotency, distributivity and absorption
 2.3.3 Other union operations

3) Drastic sum A • B
 A ( x), when  B ( x) = 0

x  X ,  A 
•
B ( x ) =  B ( x), when  A ( x) = 0
 1, for others


4) Hamacher’s sum AB

 A ( x) +  B ( x) − (2 −  )  A ( x)  B ( x)
x  X ,  A B ( x) = ,  0
1 − (1 −  )  A ( x)  B ( x)
2.4 Fuzzy intersection

2.4.1 Axioms for intersection function

I:[0,1]  [0,1] → [0,1]
 AB ( x) = I[ A ( x), B ( x)]
(Axiom I1) I(1, 1) = 1, I(1, 0) = 0, I(0, 1) = 0, I(0, 0) = 0
(Axiom I2) I(a, b) = I(b, a), Commutativity holds.
(Axiom I3) If a  a’ and b  b’, I(a, b)  I(a’, b’),
Function I is a monotonic function.
(Axiom I4) I(I(a, b), c) = I(a, I(b, c)), Associativity holds.
(Axiom I5) I is a continuous function
(Axiom I6) I(a, a) = a, I is idempotency.
2.4 Fuzzy intersection

2.4.2 Examples of intersection

◼ standard fuzzy intersection
I[A(x), B(x)] = Min[A(x), B(x)], or
AB(x) = Min[A(x), B(x)]

AB
1

X
2.4 Fuzzy intersection

Yager intersection function

I w (a, b) = 1 − Min [1, ((1 − a) w + (1 − b) w )1/ w ], w  (0, )
B 0 0.25 0.5
a
1 0 0.25 0.5 I1(a,b) =1-Min[1, 2-a-b]

0.75 0 0 0.25

0.25 0 0 0 w=1

0 0.25 0.5
a B

1 0 0.25 0.5 I2(a,b) = 1-Min[1, (1 − a ) 2 + (1 − b ) 2 ]

0.75 0 0.21 0.44

0.25 0 0 0.1 w=2

0 0.25 0.5
a
1 0 0.25 0.5 I(a,b) = Min[ a, b]

0.75 0 0.25 0.5

0.25 0 0.25 0.25 w→

 2.4.3 Other intersection operations

1) Algebraic product A • B (Probabilistic product)

xX, A•B (x) = A(x) • B(x)
◼ commutativity, associativity, identity and De Morgan’s law

2) Bounded product A • B(Bold intersection)

x  X ,  A • B ( x) = Max[0,  A ( x) +  B ( x) − 1]

◼ commutativity, associativity, identity, and De Morgan’s Law

◼ A •  = , A • A =
◼ not idempotency, distributivity and absorption
 2.4.3 Other intersection operations

3) Drastic product A • B
  A ( x), when  A ( x) = 1

 A • B ( x ) =   B ( x), when  B ( x) = 1
0, when  ( x),  ( x)  1
 A B

4) Hamacher’s product AB

 A ( x)  B ( x)
 A B ( x) = ,  0
 + (1 −  )(  A ( x) +  B ( x) −  A ( x)  B ( x))
2.5 Other operations in fuzzy set

2.5.1 Disjunctive sum

A  B = ( A  B )  ( A  B)

A B

Fig 2.10 Disjunctive sum of two crisp sets

 2.5 Other operations in fuzzy set

Simple disjunctive sum

 A (x ) = 1 - A(x) ,  B (x) = 1 - B(x)

 A B ( x) = Min[  A ( x), 1 −  B ( x )]

 A B ( x) = Min[1 −  A ( x) ,  B (x )]

A  B = ( A  B )  ( A  B ), then
 A  B ( x ) = Max{Min[  A ( x ), 1 −  B ( x )] , Min[1 −  A ( x ) , B (x )]}
2.5 Other operations in fuzzy set

Simple disjunctive sum(2)

ex) A = {(x1, 0.2), (x2, 0.7), (x3, 1), (x4, 0)}
B = {(x1, 0.5), (x2, 0.3), (x3, 1), (x4, 0.1)}

A = {(x1, 0.8), (x2, 0.3), (x3, 0), (x4, 1)}

B = {(x1, 0.5), (x2, 0.7), (x3, 0), (x4, 0.9)}

A  B = {(x1, 0.2), (x2, 0.7), (x3, 0), (x4, 0)}

A  B = {(x1, 0.5), (x2, 0.3), (x3, 0), (x4, 0.1)}

A  B = ( A  B )  ( A  B) = {(x1, 0.5), (x2, 0.7), (x3, 0), (x4, 0.1)}

2.5 Other operations in fuzzy set

Simple disjunctive sum(3)

Set A
1.0 Set B
1.0 Set A  B
0.9
0.8
0.7
0.7
0.6
0.5
0.5
0.4
0.3
0.3
0.2
0.2 0.1
0.1
0
x1 x2 x3 x4

Fig 2.11 Example of simple disjunctive sum

2.5 Other operations in fuzzy set

(Exclusive or) disjoint sum

 AB ( x) =  A ( x) −  B ( x)

Set A
1.0 Set B
1.0 Set A  B shaded area
0.9
0.8
0.7
0.7
0.6
0.5
0.5
0.4
0.3
0.3
0.2
0.2 0.1
0.1
0
x1 x2 x3 x4

Fig 2.12 Example of disjoint sum (exclusive OR sum)

2.5 Other operations in fuzzy set

(Exclusive or) disjoint sum

 AB ( x) =  A ( x) −  B ( x)

Set A
1.0 Set B
1.0 Set A  B shaded area
0.9
0.8
0.7
0.6
0.7 A = {(x1, 0.2), (x2, 0.7), (x3, 1), (x4, 0)}
0.5 B = {(x1, 0.5), (x2, 0.3), (x3, 1), (x4, 0.1)}
A△B = {(x1, 0.3), (x2, 0.4), (x3, 0), (x4, 0.1)}
0.5
0.4
0.3
0.3
0.2
0.2 0.1
0.1
0
x1 x2 x3 x4

Fig 2.12 Example of disjoint sum (exclusive OR sum)

2.5 Other operations in fuzzy set

2.5.2 Difference in fuzzy set

◼ Difference in crisp set

A− B = A B
A B

Fig 2.13 difference A – B

2.5 Other operations in fuzzy set

Simple difference
 A− B =  A B = Min [ A ( x), 1 −  B ( x)]

ex)
A = {(x1, 0.2), (x2, 0.7), (x3, 1), (x4, 0)}
B = {(x1, 0.5), (x2, 0.3), (x3, 1), (x4, 0.1)}

B = {(x1, 0.5), (x2, 0.7), (x3, 0), (x4, 0.9)}

A – B = A  B = {(x1, 0.2), (x2, 0.7), (x3, 0), (x4, 0)}

2.5 Other operations in fuzzy set

Simple difference(2)

1 Set A

Set B
A
0.7 Simple difference A-B :
shaded area
B
0.5
0.7
0.3
0.2
0.1 0.2

x1 x2 x3 x4

Fig 2.14 simple difference A – B

2.5 Other operations in fuzzy set

Bounded difference
AB(x) = Max[0, A(x) - B(x)]
Set A
Set B
1
Bounded difference : shaded area
A
0.7
B 0.4
0.5

0.3
0.2
A  B = {(x1, 0), (x2, 0.4), (x3, 0), (x4, 0)}
0.1

x1 x2 x3 x4

Fig 2.15 bounded difference A  B

2.5.3 Distance in fuzzy set

Hamming distance
n
d(A, B) = 
i =1, xiX
A ( xi ) −  B ( xi )

1. d(A, B)  0
2. d(A, B) = d(B, A)
3. d(A, C)  d(A, B) + d(B, C)
4. d(A, A) = 0

ex) A = {(x1, 0.4), (x2, 0.8), (x3, 1), (x4, 0)}

B = {(x1, 0.4), (x2, 0.3), (x3, 0), (x4, 0)}
d(A, B) = |0| + |0.5| + |1| + |0| = 1.5
2.5.3 Distance in fuzzy set

Hamming distance : distance and difference of fuzzy set

A(x) B(x)
1 1
B
A

x x
A(x) A(x)
B(x) A B(x) A
1 1
B B

x x
distance between A, B difference A- B
2.5.3 Distance in fuzzy set

Euclidean distance
n
e( A, B) =  A
( 
i =1
( x ) −  B ( x )) 2

ex)
e( A, B) = 02 + 0.52 + 12 + 02 = 1.25 = 1.12

Minkowski distance
1/ w
 w
d w ( A, B) =    A ( x) −  B ( x)  , w  [1, ]
 xX 
 2.5.4 Cartesian product of fuzzy set

Power of fuzzy set

 A ( x) = [ A ( x)]2 , x  X
2

 A ( x) = [ A ( x)]m , x  X
m

Cartesian product
 A ( x),  A ( x),  ,  A (x) as membership functions of A1, A2,, An
1 2 n

for x1  A1 , x2  A2 ,  , xn  An .

 A  A  A ( x1 , x2 , , xn ) = Min [  A ( x1 ), ,  A ( xn )]
1 2 n 1 n
 2.6 t-norms and t-conorms

2.6.1 Definitions for t-norms and t-conorms

t-norm
T : [0,1][0,1]→[0,1]
x, y, x’, y’, z  [0,1]
i) T(x, 0) = 0, T(x, 1) = x : boundary condition
ii) T(x, y) = T(y, x) : commutativity
iii) (x  x’, y  y’) → T(x, y)  T(x’, y’) : monotonicity
iv) T(T(x, y), z) = T(x, T(y, z)) : associativity

1) intersection operator (  )
2) algebraic product operator ( • )
3) bounded product operator ( • )
4) drastic product operator ( • )
2.6 t-norms and t-conorms

t-conorm (s-norm)
T : [0,1][0,1]→[0,1]
x, y, x’, y’, z  [0,1]
i) T(x, 0) = 0, T(x, 1) = 1 : boundary condition
ii) T(x, y) = T(y, x) : commutativity
iii) (x  x’, y  y’) → T(x, y)  T(x’, y’) : monotonicity
iv) T(T(x, y), z) = T(x, T(y, z)) : associativity

1) union operator (  )
2) algebraic sum operator ( +̂)
3) bounded sum operator (  )
4) drastic sum operator (• )
5) disjoint sum operator (  )
 2.6 t-norms and t-conorms

Ex)
a)  : minimum
Instead of *, if  is applied
x1=x
Since this operator meets the previous conditions, it is a t-norm.

b)  : maximum
If  is applied instead of *,
x0=x
then this becomes a t-conorm.
2.6 t-norms and t-conorms

2.6.2 Duality of t-norms and t-conorms

⊥ ( x, y ) = 1 − T ( x, y ) x ⊥ y : t − norm
x T y : t − conorm
x = 1− x
y = 1− y

xTy = 1 − T( x, y )

x ⊥ y = xTy
x Ty = x ⊥ y by De Morgane' s Law