Upload
nguyenxuyen
View
221
Download
0
Embed Size (px)
Citation preview
1
FUZZY SETS
Precision vs. Relevancy
31
is approachingyour head
at 45.3 m/sec.
A 1500 Kg massis approaching
your headat 45.3 m/s. OUT!!
LOOKOUT!
2
Introduction
How to simplify very complex systems?Allow some degree of uncertainty in theirdescription!
How to deal mathematically with uncertainty?Using probabilistic theory (stochastic).Using the theory of fuzzy sets (non-stochastic).
Proposed in 1965 by Lotfi Zadeh (Fuzzy Sets,Information Control, 8, pp. 338-353).Imprecision or vagueness in natural language doesnot imply a loss of accuracy or meaningfulness!
32
Examples
Give travel directions in terms of city blocks OR inmeters?The day is sunny OR the sky is covered by 5% of clouds?
If the sky is covered by 10% of clouds is still sunny?And 25%?And 50%?Where to draw the line from sunny to not sunny?Member and not member or membership degree?
33
3
Probability vs. Possibility
Event u: Hans ate X eggs for breakfast.Probability distribution: PX(u)Possibility distribution: X(u)
34
u 1 2 3 4 5 6 7 8
PX(u) 0.1 0.8 0.1 0 0 0 0 0
X(u) 1 1 1 1 0.8 0.6 0.4 0.2
You’re lost in the Outback; Dying of Thirst
You Come Upon Two Bottles Containing Liquid
Which One Will You Choose?
How Will You Process the Information?
35
Probability vs. Fuzzy Set Membership
4
36
Potable?Probability
= 0.91
OR
Potable?Membership
= 0.91
Might Taste Funky,but shouldn’t kill you
Probability vs. Fuzzy Set Membership
Applications of fuzzy sets
Fuzzy mathematics (measures, relations, topology,etc.)Fuzzy logic and AI (approximate reasoning, expertsystems, etc.)Fuzzy systems
Fuzzy modelingFuzzy control, etc.
Fuzzy decision makingMulti-criteria optimizationOptimization techniques
37
5
Classical set theory
Set: collection of objects with a common property.Examples:
Set of basic colors:A = {red, green, blue}
Set of positive integers:
38
{ | 0}A x x
A line in 3:A = {(x,y,z) | ax + by + cz +d = 0}
Representation of sets
Enumeration of elements: A = {x1, x2,..., xn}Definition by property P: A = {x X | P(x)}Characteristic function A(x): X {0,1}
39
1, if is member of( )
0, if is not member ofA
x Ax
x A
( ) mod 2A x xExample:
Set of odd numbers:
6
Set operations
Intersection: C = A BC contains elements that belong to A and BCharacteristic function: C = min( A , B) = A · B
Union: C = A BC contains elements that belong to A or to BCharacteristic function: C = max( A , B)
Complement: C =C contains elements that do not belong to ACharacteristic function: C = 1 – A
40
Fuzzy sets
Represent uncertain (vague, ambiguous, etc.)knowledge in the form of propositions, rules, etc.
Propositions:
expensive cars,cloudy sky,...
Rules (decisions):Want to buy a big and new house for a low price.If the temperature is low, then increase the heating.…
41
7
Classical set
Example: set of old people A = {age | age 70}
42
A
50 600
70 80 90 100
0.5
1
age [years]
Logic propositions
“Nick is old” ... true or falseNick’s age:
ageNick = 70, A(70) = 1 (true)ageNick = 69.9, A(69.9) = 0 (false)
43
A
50 600
70 80 90 100
0.5
1
age [years]
8
Fuzzy set
Graded membership, element belongs to a set to acertain degree.
44
A
50 600
70 80 90 100
0.5
1
mem
bers
hip
grad
e
age [years]
Mem
bers
hip
grad
e
Fuzzy proposition
“Nick is old”... degree of truthageNick = 70, A(70) = 0.5ageNick = 69.9, A(69.9) = 0.49ageNick = 90, A(90) = 1
45
A
50 600
70 80 90 100
0.5
1
mem
bers
hip
grad
e
age [years]
9
Context dependent
46
h [cm]
A
0170 180 190
0.5
1
mem
bers
hip
grad
e
tall in China tall in USA tall in NBA
Typical linguistic values
47
20 400
60 80 100
1
mem
bers
hip
grad
e
young middle age old
age [years]
10
Support of a fuzzy set
supp(A) = {x X | A(x) > 0}
48
xsupp( )A
A
Core (nucleous, kernel)
core(A) = {x X | A(x) = 1}
49
x
A
core( )A
11
-cut of a fuzzy set
Crisp set: A = { x X | A(x) }Strong -cut: A = { x X | A(x) > }
50
xA
A
Resolution principle
Every fuzzy set A can be uniquely represented as acollection of -level sets according to
Resolution principle implies that fuzzy set theory is ageneralization of classical set theory, and that itsresults can be represented in terms of classical settheory.
51
[0,1]( ) sup [ ( )]A Ax x
12
Resolution principle
52
x
1
0
A
Other properties
Height of a fuzzy set: hgt(A) = sup A(x), x XFuzzy set is normal(ized) when hgt(A) = 1.A fuzzy set A is convex iff x,y X and [0,1]:
A( x + (1 – ) y) min( A(x), A(y))
53
x
A B
13
Other properties (2)
Fuzzy singleton: single point x X where A(x) = 1.Fuzzy number: fuzzy set in that is normal andconvex.Two fuzzy sets are equal (A = B) iff:
x X, A(x) = B(x)A is a subset of B iff:
x X, A(x) B(x)
54
Non-convex fuzzy sets
Example: car insurance risk
55
lifetimeage [years]
14
Representation of fuzzy sets
Discrete Universe of Discourse:Point-wise as a list of membership/element pairs:
A = A(x1)/x1+...+ A(xn)/xn = i A(xi)/xi
A = { A(x1)/x1,..., A(xn)/xn} = { A(xi)/xi | xi X}
As a list of -level/ -cut pairs:A = { 1/A 1
, ..., n/A n} = { i/A i
| i [0,1]}
56
Representation of fuzzy sets
Continuous Universe of Discourse:
A = X A(x)/x
Analytical formula:
Various possible notations:
A(x), A(x), A, a, etc.
57
2
1( ) ,1A x x
x
15
Examples
Discrete universeFuzzy set A =“sensible number of children”.
number of children: X = {0, 1, 2, 3, 4, 5, 6}A = 0.1/0 + 0.3/1 + 0.7/2 + 1/3 + 0.6/4 + 0.2/5 + 0.1/6
Fuzzy 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)}
58
Examples
Continuous universeFuzzy set B = “about 50 years old”
X = + (set of positive real numbers)
B = {(x, B(x)) | x X}
59
41( )
50110
B xx
Mem
bers
hip
Gra
des
Age
16
Complement of a fuzzy set
c: [0,1] [0,1]; A(x c( A(x))
Fundamental axioms1. Boundary conditions - c behaves as the ordinary
complementc(0) = 1; c(1) = 0
2. Monotonic non-increasinga,b [0,1], if a < b, then c(a) c(b)
60
Complement of a fuzzy set
Other axioms:
c is a continuous function.c is involutive, which means that
c(c(a)) = a, a [0,1]
61
17
Complement of a fuzzy set
Equilibrium pointc(a) = a = ec, a [0,1]
Each complement has at most oneequilibrium.If c is a continuous fuzzy complement, it hasone equilibrium point.
62
Examples of fuzzy complements
Satisfying only fundamental axioms:
63
tata
acif,0if,1
)(
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.2
0.4
0.6
0.8
1
Fuzzy complement of threshold type: t=0.3
18
Examples of fuzzy complements
Satisfying fundamental axioms and continuity:
64
aac cos121)(
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.2
0.4
0.6
0.8
1
Continuous fuzzy complement (non involutive)
Examples of fuzzy complements
Sugeno complement:
65
1( ) , ] 1, ]1
ac aa
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1Sugeno fuzzy complement, lambda = -0.9, -0.5, 0, 2, 10
19
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1Yager fuzzy complement, w = 0.2, 0.5, 1, 2, 10
Examples of fuzzy complement
Yager complement:
66
],0],1)(1
waacww
w
Representation of complement
(x) = 1 – A(x)
67
1
x
A
20
0 20 40 60 80 1000
0.5
1
x
0 20 40 60 80 1000
0.5
1
x
0 20 40 60 80 1000
0.5
1
x
A(x)
1- A(x)
A(x)
=-0.9=10
A(x)
w=0.2w=5
Representation of complement
68
Intersection of fuzzy sets
i: [0,1] [0,1] [0,1];A B(x i( A(x), B(x))
Fundamental axioms: triangular norm or t-norm1. Boundary conditions - i behaves as the classical
intersectioni(1,1) = 1;i(0,1) = i(1,0) = i(0,0) = 0
2. Commutativityi(a,b) = i(b,a)
69
21
Intersection of fuzzy sets
3. MonotonicityIf a a’ and b b’, then i(a,b) i(a’,b’)
4. Associativityi(i(a,b),c) = i(a,i(b,c))
Other axioms:i is a continuous function.i(a,a) = a (idempotent).
70
Examples of fuzzy conjunctions
Zadeh
A B(x) = min( A(x), B(x))Probabilistic
A B(x) = A(x) B(x)Lukaziewicz
A B(x) = max(0, A(x) B(x) 1)
71
22
Intersection of fuzzy sets
A B(x) = min( A(x), B(x))
72
x
BA
Intersection of fuzzy sets
73
0 20 40 60 80 1000
0.2
0.4
0.6
0.8
1
x
ABmin(A,B)A*Bmax(0,A+B-1)
23
Yager t-norm
Example of week and strong intersections:
74
1( , ) 1 min 1, 1 1 , ]0, ]
ww wwi a b a b w
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.2
0.4
0.6
0.8
1
Yager fuzzy intersection, w = 1.5, 2, 5
Union of fuzzy sets
u: [0,1] [0,1] [0,1];A B(x u( A(x), B(x))
Fundamental axioms: triangular co-norm or s-norm1. Boundary conditions - u behaves as the classical
unionu(0,0) = 0;u(0,1) = u(1,0) = u(1,1) = 1
2. Commutativityu(a,b) = u(b,a)
75
24
Union of fuzzy sets
3. MonotonicityIf a a’ and b b’, then u(a,b) u(a’,b’)
4. Associativityu(u(a,b),c) = u(a,u(b,c))
Other axioms:u is a continuous function.u(a,a) = a (idempotent).
76
Examples of fuzzy disjunctions
Zadeh
A B(x) = max( A(x), B(x))Probabilistic
A B(x) = A(x) B(x) A(x) B(x)Lukasiewicz
A B(x) = min(1, A(x) B(x))
77
25
Union of fuzzy sets
A B(x) = max( A(x), B(x))
78
x
BA
Union of fuzzy sets
79
0 20 40 60 80 1000
0.2
0.4
0.6
0.8
1
x
ABmax(A,B)A+B-A*Bmin(1,A+B)
26
Yager s-norm
Example of week and strong disjunctions:
80
1( , ) min 1, , ]0, ]
ww wwu a b a b w
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.2
0.4
0.6
0.8
1
Yager fuzzy union, w = 2.5, 5, 10
General aggregation operations
h: [0,1]n [0,1];A(x h( A1
(x),..., An(x))
Axioms1. Boundary conditions
h(0,...,0) = 0h(1,...,1) = 1
2. Monotonic non-decreasingFor any pair ai, bi [0,1], iIf ai bi then h(ai) h(bi)
81
27
General aggregation operations
Other axioms:
h is a continuous function.h is a symmetric function in all its arguments:
h(ai) = h(ap(i))for any permutation p on
82
Averaging operations
When all the four axioms hold:
Operator covering this range: Generalized mean
83
1 1 1min( , , ) ( , , ) max( , , )n n na a h a a a a
1
11( , , ) n
n
a ah a a
n
28
Generalized mean
Typical cases:Lower bound:Geometric mean:
Harmonic mean:
Arithmetic mean:Upper bound:
84
1
1
1 1n
nh
a a
1max( , , )nh a a
1min( , , )nh a a
0 1 2( )nnh a a a
11
na ahn
Fuzzy aggregation operations
85
min max umax
Generalized mean
Parametric t-norms Parametric s-norms
Intersection operators Averaging operators Union operators
imin
29
Membership functions (MF)
Triangular MF:
Trapezoidal MF:
Gaussian MF:
Generalized bell MF:
86
( ; , , ) max min , ,0x a c xTr x a b cb a c b
( ; , , , ) max min ,1, ,0x a d xTp x a b c db a d c
212( ; , , )
x c
Gs x a b c e
2
1( ; , , )1
aBell x a b cx c
b
Membership functions
87
0 20 40 60 80 1000
0.5
1
(a) Triangular MF
0 20 40 60 80 1000
0.5
1
(b) Trapezoidal MF
0 20 40 60 80 1000
0.5
1
(c) Gaussian MF
0 20 40 60 80 1000
0.5
1
(d) Generalized Bell MF
30
Left-right MF
88
,( ; , , )
,
L
R
c xF x cLR x c
x cF x c
c = 65= 60= 10
c = 25= 10= 40
0 50 1000
0.2
0.4
0.6
0.8
1
Mem
bers
hip
Gra
des
0 50 1000
0.2
0.4
0.6
0.8
1
Mem
bers
hip
Gra
des
Two-dimensional fuzzy sets
89
( , ) ( , ) ( , )A AX YA x y x y x y X Y
x y
31
2-D membership functions
90
-100
10
-100
100
0.5
1
X
(a) z = m in(trap(x ), trap(y))
Y -100
10
-100
100
0.5
1
X
(b) z = m ax (trap(x), trap(y ))
Y
-100
10
-100
100
0.5
1
X
(c) z = m in(bell(x ), bell(y))
Y -100
10
-100
100
0.5
1
X
(d) z = m ax (bell(x), bell(y ))
Y
Compound fuzzy propositions
Small = Short and Light (conjunction)
91
( , ) ( ) ( )Small Short Lighth w h w
LONG
LIG
HT
HEA
VY
Weight(w)
Height(h)
Thin
BigCompact
Small
32
Cylindrical extension
Cylindrical extension of fuzzy set A in X into Y resultsin a two-dimensional fuzzy set in X Y, given by
92
0
0.5
1
X
(a) Base Fuzzy Set A
0
0.5
1
X
(b) Cylindrical Extension of A
y
cext ( ) ( ) /( , ) ( ) ( , ) ( , )y A AX YA x x y x x y x y X Y
Projection
93
0
0.5
1
X
(a) A Two-dimensional MF
Y
0
0.5
1
X
(b) Projection onto X
Y
0
0.5
1
X
(c) Projection onto Y
Y
( , )R x y ( ) max ( , )A Ryx x y ( ) max ( , )B Rx
y x y
33
Cartesian product
Cartesian product of fuzzy sets A and B is a fuzzy set inthe product space X Y with membership
Cartesian co-product of fuzzy sets A and B is a fuzzyset in the product space X Y with membership
94
( , ) min( ( ), ( ))A B A Bx y x y
( , ) max( ( ), ( ))A B A Bx y x y
Cartesian product
95
-10 0 100
0.5
1
X
(X)
(a) trap(x)
-10 0 100
0.5
1
Y
(Y)
(b) trap(y)
-100
10-10
010
0
0.5
1
X
(c) z = min(trap(x), trap(y))
Y -100
10-10
010
0
0.5
1
X
(d) z = max(trap(x), trap(y))
Y
34
Classical relations
Classical relation R(X1, X2,..., Xn) is a subset of theCartesian product:
Characteristic function:
96
1 2 1 2, , , n nX X X X X XR
1 21 2
1, iff , , ,, , ,
0, otherwisen
n
x x xx x xR
R
Example
X = {English, French}Y = {dollar, pound, euro}Z = {USA, France, Canada, Britain, Germany}
R(X, Y, Z) = {(English, dollar, USA),(French, euro, France), (English, dollar,Canada),
(French, dollar, Canada), (English, pound,Britain)}
97
35
Matrix representation
98
USA Fra Can Brit GerDollar 1 0 1 0 0Pound 0 0 0 1 0
Euro 0 0 0 0 0 English
USA Fra Can Brit GerDollar 0 0 1 0 0Pound 0 0 0 0 0
Euro 0 1 0 0 0 French
Fuzzy relation
Fuzzy relation:
R: X1 X2 ... Xn [0,1]
Each tuple (x1, x2,..., xn) has a degree of membership.Fuzzy relation can be represented by an n-dimensionalmembership function (continuous space) or a matrix(discrete space).Examples:
x is close to yx and y are similarx and y are related (dependent)
99
36
Discrete examples
Relation R “very far” between X = {New York, Lisbon}and Y = {New York, Beijing, London}:R(x,y) = 0/(NY,NY) + 1/(NY,Beijing) +
0.6/(NY,London) + 0.5/(Lisbon,NY) +0.8/(Lisbon,Beijing) + 0.1/(Lisbon,London)
Relation: “is an important trade partner of”
100
Holland Germany USA JapanHolland 1 0,9 0,5 0,2
Germany 0,3 1 0,4 0,2USA 0,3 0,4 1 0,7
Japan 0,6 0,8 0,9 1
Continuous example
R: x y (“x is approximately equal to y”)
101
2( )( , ) x yx y e
-1-0.5
00.5
1
-1
-0.5
00.5
10
0.5
1
xy
37
Composition of relations
R(X,Z) = P(X,Y) Q(Y,Z)
Conditions:(x,z) R iff exists y Y such that(x,y) P and (y,z) Q.
Max-min composition
102
( , ) max min ( , ), ( , )y Y
x z x y y zP Q P Q
Properties
Associativity:
Distributivity over union:
Weak distributivity over intersection:
Monotonicity:
103
( ) ( )R S T R S T
( ) ( ) ( )R S T R S R T
( ) ( ) ( )R S T R S R T
( ) ( )S T R S R T
38
Other compositions
Max-prod composition
Max-t composition
104
( , ) max ( , ) ( , )y Y
x z x y y zP Q P Q
( , ) max ( , ), ( , )y Y
x z t x y y zP Q P Q
Example
105
YZa
b
c
d
1
2
Q
P
0.70.5
1
1
0.40.3
0.6
0.8
1
0.9
39
Example
Composition R = P Q
Composition R = P Q?
106
x z R(x,z)a 1 0.6a 2 0.7b 1 0.6b 2 0.8c 2 1d 2 0.4
Matrix notation examples
107
0.3 0.5 0.8 0.9 0.5 0.7 0.70 0.7 1 0.3 0.2 0 0.9
0.4 0.6 0.5 1 0 0.5 0.5
0.3 0.5 0.8 0.9 0.5 0.7 0.70 0.7 1 0.3 0.2 0 0.9
0.4 0.6 0.5 1 0 0.5 0.5
0.8 0.3 0.5 0.51 0.2 0.5 0.7
0.5 0.4 0.5 0.6
0.8 0.15 0.4 0.451 0.14 0.5 0.63
0.5 0.2 0.28 0.54
40
Relations on the same universe
Let R be a relation defined on U U, then it is called:Reflexive, if u U, the pair (u,u) RAnti-reflexive, if u U, (u,u) RSymmetric, if u,v U, if (u,v) R, then (v,u) RtooAnti-symmetric, if u,v U, if (u,v) and (v,u) R,then u = vTransitive, if u,v,w U , if (u,v) and (v,w) R, then(u,w) R too.
108
Examples
R is an equivalence relation if it is reflexive, symmetricand transitive.R is a partial order relation if it is reflexive, anti-symmetric and transitive.R is a total order relation if R is a partial orderrelation, and u, v U, either (u,v) or (v,u) R.Examples:
The subset relation on sets ( ) is a partial orderrelation.The relation on is a total order relation.
109