Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 1
Chapter 2: FUZZY SETSIntroduction (2.1)
Basic Definitions &Terminology (2.2)
Set-theoretic Operations (2.3)
Membership Function (MF) Formulation & Parameterization (2.4)
More on Fuzzy Union, Intersection & Complement (2.5)
Jyh-Shing Roger Jang et al., Neuro-Fuzzy and Soft Computing: A Computational Approach to Learning and Machine Intelligence, First Edition, Prentice Hall, 1997
Introduction (2.1)
Sets with fuzzy boundaries
A = Set of tall people
Heights5’10’’
1.0
Crisp set A
Membershipfunction
Heights5’10’’ 6’2’’
.5
.9
Fuzzy set A1.0
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 2
Membership Functions (MFs)
Characteristics of MFs:Subjective measuresNot probability functions
MFs
Heights
.8
“tall” in Asia
.5 “tall” in the US
5’10’’.1 “tall” in NBA
Introduction (2.1) (cont.)
Basic definitions & Terminology (2.2)
Formal definition:A fuzzy set A in X is expressed as a set of ordered
pairs:
A x x x XA= ∈{( , ( ))| }µ
Universe oruniverse of discourse
Fuzzy set Membershipfunction(MF)
A fuzzy set is totally characterized by aA fuzzy set is totally characterized by amembership function (MF).membership function (MF).
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 3
Fuzzy Sets with Discrete UniversesFuzzy set C = “desirable city to live in”X = {SF, Boston, LA} (discrete and non-ordered)C = {(SF, 0.9), (Boston, 0.8), (LA, 0.6)}(subjective membership values!)
Fuzzy set A = “sensible number of children”X = {0, 1, 2, 3, 4, 5, 6} (discrete universe)A = {(0, .1), (1, .3), (2, .7), (3, 1), (4, .6), (5, .2), (6, .1)}(subjective membership values!)
Basic definitions & Terminology (2.2) (cont.)
Fuzzy Sets with Cont. Universes
Fuzzy set B = “about 50 years old”X = Set of positive real numbers (continuous)B = {(x, µB(x)) | x in X}
µ B x x( ) =
+−
1
1 5010
2
Basic definitions & Terminology (2.2) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 4
Alternative Notation
A fuzzy set A can be alternatively denoted as follows:
A x xAX
= ∫ µ ( ) /X is continuous
A x xAx X
i ii
=∈
∑ µ ( ) /X is discrete
Note that Σ and integral signs stand for the union of membership grades; “/” stands for a marker and does not imply division.
Basic definitions & Terminology (2.2) (cont.)
Fuzzy Partition
Fuzzy partitions formed by the linguistic values “young”, “middle aged”, and “old”:
Basic definitions & Terminology (2.2) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 5
Support(A) = {x ∈ X | µA(x) > 0}
Core(A) = {x ∈ X | µA(x) = 1}
Normality: core(A) ≠ ∅ ⇒ A is a normal fuzzy set
Crossover(A) = {x ∈ X | µA(x) = 0.5}
α - cut: Aα = {x ∈ X | µA(x) ≥ α}
Strong α - cut: A’α = {x ∈ X | µA(x) > α}
Basic definitions & Terminology (2.2) (cont.)
Crossover points
Support
α - cut
Core
MF
X0
.5
1
α
MF Terminology
Basic definitions & Terminology (2.2) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 6
Convexity of Fuzzy Sets
A fuzzy set A is convex if for any λ in [0, 1],
µ λ λ µ µA A Ax x x x( ( ) ) min( ( ), ( ))1 2 1 21+ − ≥Alternatively, A is convex if all its α-cuts are convex.
Basic definitions & Terminology (2.2) (cont.)
Fuzzy numbers: a fuzzy number A is a fuzzy set in IR that satisfies normality & convexity
Bandwidths: for a normal & convex set, the bandwidth is the distance between two unique crossover points
Width(A) = |x2 – x1|With µA(x1) = µA(x2) = 0.5
Symmetry: a fuzzy set A is symmetric if its MF is symmetric around a certain point x = c, namely
µA(x + c) = µA(c – x) ∀x ∈ X
Basic definitions & Terminology (2.2) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 7
Open left, open right, closed:
0)x(lim)x(lim set A fuzzy closed
1)x(lim and 0)x(lim set A fuzzyright open
0)x(limand1)x(limset A fuzzyleft open
AxA-x
AxA-x
AxA-x
=µ=µ⇔
=µ=µ⇔
=µ=µ⇔
+∞→∞→
+∞→∞→
+∞→∞→
Basic definitions & Terminology (2.2) (cont.)
Set-Theoretic Operations (2.3)
Subset:A B A B⊆ ⇔ ≤µ µ
C A B x x x x xc A B A B= ∪ ⇔ = = ∨µ µ µ µ µ( ) max( ( ), ( )) ( ) ( )
C A B x x x x xc A B A B= ∩ ⇔ = = ∧µ µ µ µ µ( ) min( ( ), ( )) ( ) ( )
A X A x xA A= − ⇔ = −µ µ( ) ( )1
Complement:
Union:
Intersection:
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 8
Set-Theoretic Operations (2.3) (cont.)
MF Formulation & Parameterization (2.4)
Triangular MF: trim f x a b cx ab a
c xc b
( ; , , ) max min , ,=−−
−−
0
trapm f x a b c dx ab a
d xd c
( ; , , , ) max min , , ,=−−
−−
1 0
gbellm f x a b cx c
b
b( ; , , ) =+
−
1
12
2cx21
e),c;x(gaussmf
σ−
−=σ
Trapezoidal MF:
Gaussian MF:
Generalized bell MF:
MFs of One Dimension
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 9
MF Formulation & Parameterization (2.4) (cont.)
Change of parameters in the generalized bell MF
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 10
Physical meaning of parameters in a generalized bell MF
MF Formulation & Parameterization (2.4) (cont.)
Gaussian MFs and bell MFs achieve smoothness, they are unable to specify asymmetric Mfs which are important in many applications
Asymmetric & close MFs can be synthesized using either the absolute difference or the product of two sigmoidal functions
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 11
Sigmoidal MF: )cx(ae11)c,a;x(sigmf −−+
=
Extensions:
Productof two sig. MF
Abs. differenceof two sig. MF
MF Formulation & Parameterization (2.4) (cont.)
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
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 12
Left –Right (LR) MF:
L R x cF
c xx c
Fx c
x c
L
R
( ; , , ),
,α β
α
β
=
−
<
−
≥
Example: F x xL ( ) max( , )= −0 1 2 F x xR ( ) exp( )= −3
c=65a=60b=10
c=25a=10b=40
MF Formulation & Parameterization (2.4) (cont.)
The list of MFs introduced in this section is by no means exhaustive
Other specialized MFs can be created for specific applications if necessary
Any type of continuous probability distribution functions can be used as an MF
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 13
MFs of two dimensions
In this case, there are two inputs assigned to an MF: this MF is a twp dimensional MF. A one input MF is called ordinary MF
Extension of a one-dimensional MF to a two-dimensional MF via cylindrical extensions
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
∫ µ=Y*X
A )y,x(|)x()A(C
MF Formulation & Parameterization (2.4) (cont.)
Cylindrical extensionBase set A Cylindrical Ext. of A
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 14
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:
and
respectively.
x|)y,x(maxRX
RyX ∫
µ=
∫
µ=
YRxY
y|)y,x(maxR
MF Formulation & Parameterization (2.4) (cont.)
Two-dimensionalMF
Projectiononto X
Projectiononto Y
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 15
Composite & non-composite MFs
Suppose that the fuzzy A = “(x,y) is near (3,4)” is defined by:
This two-dimensional MF is compositeThe fuzzy set A is composed of two statements:
“x is near 3” & “y is near 4”
( )
)1,4;y(G*)2,3;x(G 1
4yexp2
3xexp
4y2
3xexp)y,x(
22
22
A
=
−−
−−=
−−
−−=µ
MF Formulation & Parameterization (2.4) (cont.)
These two statements are respectively defined as:µ near 3 (x) = G(x;3,2) & µ near 4 (x) = G(y;4,1)
If a fuzzy set is defined by:
it is non-composite.
A composite two-dimensional MF is usually the result of two statements joined by the AND or OR connectives.
5.2A |4y||3x|11)y,x(
−−+=µ
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 16
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.
MF Formulation & Parameterization (2.4) (cont.)
Two dimensional MFs defined by the min and max operators
MF Formulation & Parameterization (2.4) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 17
More on Fuzzy Union, Intersection & Complement (2.5)
Fuzzy complement
Another way to define reasonable & consistent operations on fuzzy sets
General requirements:
Boundary: N(0)=1 and N(1) = 0Monotonicity: N(a) > N(b) if a < bInvolution: N(N(a) = a
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
Two types of fuzzy complements:
Sugeno’s complement:
(Family of fuzzy complement operators)
Yager’s complement:
N a aww w( ) ( ) /= −1 1
N a asas
( ) = −+
11 (s > -1)
(w > 0)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 18
N a asas
( ) = −+
11
Sugeno’s complement:
N a aww w( ) ( ) /= −1 1
Yager’s complement:
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
Fuzzy Intersection and Union:
The intersection of two fuzzy sets A and B is specified in general by a function
T: [0,1] * [0,1] → [0,1] with
This class of fuzzy intersection operators are called T-norm (triangular) operators.
( )
T. function the for operator binary a is * where
)x(*)x()x(),x(T)x(~
B~
ABABA µµ=µµ=µ ∩
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 19
T-norm operators satisfy:
Boundary: T(0, 0) = 0, T(a, 1) = T(1, a) = aCorrect generalization to crisp sets
Monotonicity: T(a, b) < T(c, d) if a < c and b < dA decrease of membership in A & B cannot increase a membership in A ∩ B
Commutativity: T(a, b) = T(b, a)T is indifferent to the order of fuzzy sets to be combined
Associativity: T(a, T(b, c)) = T(T(a, b), c)Intersection is independent of the order of pairwise groupings
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
T-norm (cont.)
Four examples (page 37):
Minimum: Tm(a, b) = min (a,b) = a ∧ b
Algebraic product: Ta(a, b) = ab
Bounded product: Tb(a, b) = 0 V (a + b – 1)
Drastic product: Td(a, b) =
<==
1ba, if 0,1 a if b,1b if ,a
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 20
T-norm Operator
Minimum:Tm(a, b)
Algebraicproduct:Ta(a, b)
Boundedproduct:Tb(a, b)
Drasticproduct:Td(a, b)
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
T-conorm or S-norm
The fuzzy union operator is defined by a functionS: [0,1] * [0,1] → [0,1]
wich aggregates two membership function as:
where s is called an s-norm satisfying:
Boundary: S(1, 1) = 1, S(a, 0) = S(0, a) = aMonotonicity: S(a, b) < S(c, d) if a < c and b < dCommutativity: S(a, b) = S(b, a)Associativity: S(a, S(b, c)) = S(S(a, b), c)
( ) )x()x()x(),x(S B~
ABABA µ+µ=µµ=µ ∪
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 21
T-conorm or S-norm (cont.)
Four examples (page 38):
Maximum: Sm(a, b) = max(a,b) = a V b
Algebraic sum: Sa(a, b) = a + b - ab
Bounded sum: Sb(a, b) = 1 ∧ (a + b)
Drastic sum: Sd(a, b) =
>==
0ba, if 1,0a if b,0b if ,a
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
T-conorm or S-norm
Maximum:Sm(a, b)
Algebraicsum:
Sa(a, b)
Boundedsum:
Sb(a, b)
Drasticsum:
Sd(a, b)
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
CSE 513 Soft Computing Dr. Djamel Bouchaffra
Ch.2: Fuzzy sets 22
Generalized DeMorgan’s Law
T-norms and T-conorms are duals which support the generalization of DeMorgan’slaw:
T(a, b) = N(S(N(a), N(b)))S(a, b) = N(T(N(a), N(b)))
Tm(a, b)Ta(a, b)Tb(a, b)Td(a, b)
Sm(a, b)Sa(a, b)Sb(a, b)Sd(a, b)
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
Parameterized T-norm and T-conorm
Parameterized T-norms and dual T-conorms have been proposed by several researchers:
YagerSchweizer and SklarDubois and PradeHamacherFrankSugenoDombi
More on Fuzzy Union, Intersection & Complement (2.5) (cont.)
Chapter 2: FUZZY SETSIntroduction (2.1)Introduction (2.1) (cont.)Basic definitions & Terminology (2.2)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Basic definitions & Terminology (2.2) (cont.)Set-Theoretic Operations (2.3)Set-Theoretic Operations (2.3) (cont.)MF Formulation & Parameterization (2.4)MF Formulation & Parameterization (2.4) (cont.)More on Fuzzy Union, Intersection & Complement (2.5)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)T-norm OperatorMore on Fuzzy Union, Intersection & Complement (2.5) (cont.)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)T-conorm or S-normMore on Fuzzy Union, Intersection & Complement (2.5) (cont.)More on Fuzzy Union, Intersection & Complement (2.5) (cont.)