Professional Documents
Culture Documents
Fuzzy Logic: Lofti A. Zadeh University of California, Berkeley
Fuzzy Logic: Lofti A. Zadeh University of California, Berkeley
Fuzzy Logic: Lofti A. Zadeh University of California, Berkeley
Lofti A. Zadeh
University of California, Berkeley
84 COMPUTER
bility is “almost impossible.” 1, . . ., n , is the domain of X;. Represen- membership of U in TALL or, equiva-
An important issue in fuzzy logic relates tation o f p in its canonical form requires, lently, the degree to which a numerical
to inference from qualified propositions, in general, the construction of an explana- height U satisfies the constraint induced by
especially from probability-qualified tory database and a test procedure that the relation TALL.
propositions. This issue is of central tests and aggregates the test scores Whenp is a conditional proposition, its
importance in the management of uncer- associated with the elastic constraints C1, canonical form may be expressed as “ Yis
tainty in expert systems and in the formali- . . ., c k . 6 Bif XisA,” implying thatpinduces acon-
zation of common sense reasoning. In the In more concrete terms, the canonical ditional possibility distribution of Ygiven
latter, it’s important to note the close con- form of p implies that the possibility X , written as ll(ylm. In fuzzy logic, nlylm
nection between probability-qualification distribution6 of X i s equal t o A-that is, may be defined in a variety of ways,’
and usuality-qualification and the role among which is a definition consistent
played by fuzzy quantifiers. For example, with the definition of implication in
the disposition Lakasiewicz’s LAleppho logic. In this case,
which in turn implies that the conditional possibility distribution
Swedes are blond function, n(yIx),which defines n(yim,may
Poss{X=u} = pa(u), U E U be expressed as
may be interpreted as
where pA is the membership function of A
most Swedes are blond; and Poss{X= U } is the possibility that X
may take U as its value. Thus, when the
or, equivalently, as meaning o f p is represented in the form of
Equation 1, it signifies that p induces a where
(Swede is blond) is likely, possibility distribution n, that is equal to
A , with A playing the role of an elastic n(Ylx)(u,
A Poss{X=u, Y = v J
v)=
where “likely” is a fuzzy probability that constraint on a variable X that is implicit
is numerically equal to the fuzzy quantifier i n p . In effect, the possibility distribution pA and pa denote the membership func-
“most”; or, equivalently, as of X , nx, is the set of possible values of tions of A and B, respectively; and A
X , with the understanding that possibility denotes the operator min.
usually (a Swede is blond), is a matter of degree. Viewed in this per- When p is a quantified proposition of
spective, a proposition p constrains the the form
where “usually” qualifies the proposition possible values that X can take and thus
“a Swede is blond.” defines its possibility distribution. This p 4 Q A’Sare B‘S
As alluded earlier, inference from implies that the meaning o f p is defined by
propositions of this type is a main concern (1) identifying the variable that is con- for example,
of dispositional logic. More about this strained and (2) characterizing the con-
logic will be said at a later point. straint to which the variable is subjected p A most tall men are not very fat
through its possibility distribution. Note
that Equation 1 asserts that the possibility where Q is a fuzzy quantifier and A and B
Meaning representation that Xcan take U as its value is numerically are fuzzy predicates, the constrained var-
and inference equal to the grade of membership, p A ( u ) , iable, X , is the proportion of B’s in A’s,
of u in A . with Q representing an elastic constraint
A basic idea serving as a point of depar- As an illustration, consider the propo- on X . More specifically, if Uis a finite set
ture in fuzzy logic is that a proposition p sition {U,,. . . , U,,,}, the proportion of B‘s in
in a natural or synthetic language may be A‘s is defined as the relative sigma-count
viewed as a collection of elastic con- p A John is tall
straints, C,, . . . , ck, which restrict the 1PA@,) A PB(U,)
values of a collection of variables X = in which the symbol1 should be read as ZCount @ / A ) = ’ /AA(u,)
(3)
(Xl, . . ., Xn).6In general, the constraints “denotes” or “is equal to by definition.”
as well as the variables they constrain are In this case, X = Height(John), A = j = 1,. . . , m
implicit rather than explicit in p . Viewed TALL, and the canonical form o f p reads
in this perspective, representation of the where pa(u,) and pB(u,) denote the grades
meaning of p is, in essence, a process by Height(John) is TALL of membership of uj in A and B , respec-
which the implicit constraints and vari- tively. Thus, expressed in its canonical
ables i n p are made explicit. In fuzzy logic, where the fuzzy relation TALL is in upper- form, Equation 3 may be written as
this is accomplished by representing p in case letters to underscore that it plays the
the so-called canonical form role of a constraint in the canonical form. zCount(B/A) is Q
From the canonical form, it follows that
p-+XisA which places in evidence the constrained
Poss {Height (John) = U } = ~TALL(u) variable, X, i n p and the elastic constraint,
in which A is a fuzzy predicate or, equiva- Q, to which X i s subjected. Note that X i s
lently, an n-ary fuzzy relation in U, where where /ATALL is the membership function the relative sigma-count of B in A .
U = U, x U2 x . . . x U”, and U;, i = of TALL and ~ T A L L ( u ) is the grade of The concept of a canonical form pro-
April 1988 85
1 1 expressed as the inference rule
= 1, . . . , N,may be (1) a fact that may inferring the value of a specified function
be expressed in a canonical form as “Xis f(X1, . . . ,X,,), f: U -,V , of the vari- where the fuzzy number 2 (Q1 63 Q2)
A” or (2) a rule that may be expressed in ables constrained by the knowledge base. should be read as “at least (Q163 e’).” In
a canonical form as “ Y is B,” if “X is Because of the incompleteness and impre- particular, if the fuzzy quantifiers Q1and
cision of the information resident in K B , Q2 are monotone increasing (for example,
A,.” More generally, both facts and rules
may be probability-qualified or, equiva- what we can deduce, in general, is not the when “ Q I = Q2 ! !most”), then
lently, expressed as quantified proposi- value o f f ( X l , . . . , X,,) but its possibility
tions. For example, a rule of the general distribution, n,. By employing the exten-
form “QA’s are B‘s” may be interpreted sion principle,’ it can be shown that the
as the probability-qualified proposition ( X possibility distribution function of f is and Equation 6 becomes
is B if X i s A ) is A, where A is a fuzzy prob- given by the solution of the nonlinear
ability whose denotation as a fuzzy subset program Q I A‘s are B’s (7)
of the unit interval is the same as that of Q2 ( A and B)’s are C‘s
the fuzzy quantifier Q and X i s chosen at (Q1 63 Q2)A’s are C’s
random in U. . . . ,U,>A
mtx,,. ,x,,)(U,,
Now ifp, induces a possibility distribu- ... Furthermore, if B is a subset of A , then A
tion ili;rl, J,,), where X,, . . . , X,, are and B = B , and Equation 7 reduces to the
the variables constrained by p , , then the A nkI, , x,,)( ~ 1 , ... 9 chaining rule
possibility distribution n, which
,x,,),
is induced by the totality of propositions subject to the constraint QI A‘s are E’s
in KB is given by the intersection3 of the Q2 B’s are C’s
nix1, .X,). That is, (QI 63 Q2)A’s are C’s
whereuJE UJ, i = 1 , . . . , n, and v E V.
The reduction to the solution of a non- For example,
linear program constitutes the principal
tool for inference in fuzzy logic. most students are undergraduates
or, equivalently, most undergraduates are young
Fuzzy syllogisms. A basic fuzzy syllo- most‘ students are young
gism in fuzzy logic that is of considerable
relevance to the rules of combination of where “most’” represents the product of
evidence in expert systems is the intersec- the fuzzy number “most” with itself (see
is the possibility distribution
n ( x l , ,) , , ,
, , tionlproduct syllogism-a syllogism that Figure 3).
function of ,fl, ,?,). Note that there is
, , , serves as a rule of inference for quantified What is important to observe is that the
no loss of generality in assuming that the propositions.’ This syllogism may be chaining rule expressed by Equation 8
86 COMPUTER
serves the same purpose as the chaining containment of B in A and the monotonic-
rules in Mycin, Prospector, and other ity of Ql and Q2are conditions for robust-
probability-based expert systems. How- ness in the case of the intersection/product
ever, Equation 8 is formulated in terms of syllogism.
fuzzy quantifiers rather than numerical Another basic syllogism is the conse-
probabilities or certainty factors, and it is quent conjunction sylIogism
a logical consequence of the concept of a
relative sigma-count in fuzzy logic. Fur- QI A’s are B’s (9)
thermore, the chaining rule (Equation 8) Q2 A’s are C’s
is robust in the sense that if Q1and Q2are Q A’s are ( B and C)’s
close to unity, so is their product QI C3 Q 2 .
More specifically, if QI and Q2 are where
expressed as
April 1988 87
application of fuzzy logic to the premises Mycin, Prospector, and related systems is ZCount ( C / A n B ) =
in question is both robust and composi- that the items of evidence are conditionally ECount ( C I A ) ZCount ( C / B ) d
tional. independent, given the hypothesis (and its
A more complex problem is presented complement). That is, where the factor d is expressed by
by what in Mycin and Prospector cor-
responds to the conjunctive combination t c o u n t ( A )Z c o u n t ( B )
of evidence. Stated in terms of quantified d = (12)
premises, the inference rule in question where P (E,,E2 1 H ) is the joint probabil- t c o u n t (A n B ) t c o u n t (c)
may be expressed as ity of El and E,, given the hypothesis H
and P ( E , I H ) and P (E21 H ) are the con- Inspection of Equation 12 shows that
Q1A‘s are C’s ditional probabilities of E , given H and the assumption expressed by Eqaution 11
O7B’s are C‘s E2 given H , respectively. Expressed in does not ensure the compositionality of Q.
However, it can be shown that composi-
~
Interpolation
An important problem that arises in the operation of any consistency, yf,, is defined as
rule-basedsystem is the following. Suppose the user supplies
a fact that, in its canonical form, may be expressed as “X is Yl,
A= SUP (R,l n RI)
A,” where A is a fuzzy or nonfuzzy predicate. Furthermore, sup- = (Me,, A pR,
pose that there is no conditional rule in K 6 whose antecedent
matches A exactly. The question arises: Which rules should in which pe and pe are the membership functions of R, and R,
be executed and how should their results be combined? respectivelv; U, is dgeneric element of U,;and the supremum
An approach to this problem, sketched in Reference 8, is taken over U,.
involves the use of an interpolation technique in fuzzy logic Next, we compute the overall degree of consistency, Y,, of
which requires a computation of the degree of partial match the input n-tuple ( R , , . . . , R,) with the i t h row of R, i = 1,. . . ,
between the user-suppliedfact and the rows of a decision m, by employing A (min) as the aggregation operator. Thus,
table. More specifically, suppose that upon translation into
their canonical forms, a group of propositions in K6 may be Y, = Y,I A Yi2 A ... A Ym
expressed as a fuzzy relation of the form
which implies that y , may be interpreted as a conservative
measure of agreement between the input n-tuple (R,, . . . , R,)
and the ith-row n-tuple (R,,, . . . , R,,). Then, employing y, as a
weighting coefficient, the desired expression for X,,, may be
written as a “linear” combination
X,+, = ~1 A . , . + Ym
Z1 i- A Z,,,
in which the entries are fuzzy sets; the input variables are X,,
in which + denotes the union, and Y, A z, is a fuzzy set defined
. . . , X,, with domains U,,.. . , U,;and the output variable is by
X,+,, with domain U,,,.The problem is: Given an input n-tuple
p liui
(U,+,) = Yi A pz, ( U i + r ) , i = 1, . . . v m
( R , , . . . , R,), in which R, i = 1, . . . , n, is a fuzzy subset of U,,
what is the value of X,,, expressed as a fuzzy subset of U,,,? The above approach ceases to be effective, however, when R
A possible approach to the problem is to compute for each is a sparse relation in the sense that no row of R has a high
pair (R,,, R,) the degree of consistency of the input R, with the degree of consistency with the input n-tuple. For such cases,
R,, element of R, i = 1,. . . , m , i = 1,. . . , n. The degree of a more general interpolation technique has to be employed.
88 COMPUTER
Basic rules of inference where PYOUNC and OLD are, respectively, of R and the supremum is taken over v E
the membership functions of YOUNG and V.
One distinguishing characteristic of OLD, and the universe of discourse is the
fuzzy logic is that premises and conclu- interval [0, 1001. Compositional rule.
sions in an inference rule are generally Viewed in a different perspective, the
Xis A
expressed in canonical form. This repre- entailment principle in fuzzy logic may be (X, Y ) i s R
sentation places in evidence the fact that regarded as a generalization to fuzzy sets
of the inheritance principle widely used in YisA 0 R
each premise is a constraint on a variable
and that the conclusion is an induced con- knowledge representation systems. More
specifically, if the proposition “ X i s A ” is where A o R , the composition of the
straint computed through a process of
interpreted as “X has property A ,” then unary relation A with the binary relation
constraint propagation - a process that,
the conclusion “ X is B” may be inter- R , is defined by
in general, reduces to the solution of a non-
linear program. The following briefly preted as “X has property B,” where B is
presents - without derivation - some of any superset of A. In other words, X
the basic inference rules in fuzzy logic. inherits property B if B is a superset of A .
Most of these rules can be deduced from Among other categorical rules that play The compositional rule of inference may
the basic inference rule expressed by Equa- a basic role in fuzzy logic are the follow- be viewed as a combination of the con-
tion 4. ing. In all of these rules, X , Y, Z, . . . are junctive and projection rules.
The rules of inference in fuzzy logic may variables ranging over specified universes
be classified in a variety of ways. One basic of discourse, and A , B , C, . . . are fuzzy Generalized modus ponens.
class is categoricalrules, that is, rules that predicates o r , equivalently, fuzzy
d o not contain fuzzy quantifiers. A more relations. XisA
general class is dispositionalrules, rules in Y is Cif X i s B
which one or more premises may contain, Conjunctive rule.
Y is A 0 ( T B8 C )
explicitly or implicitly, the fuzzy quantifier
“usually.” For example, the inference rule XisA
Xis B where 1 B denotes the negation of B and
known as the entailment principle: the bounded sum is defined by
XisA nB
XisA
A C B where A n B is the intersection of A and
Xis B B defined by An important feature of the generalized
modus ponens, which is not possessed by
where Xis a variable taking values in a uni- the modus ponens in binary logical sys-
verse of discourse U , and A and B are tems, is that the antecedent “Xis B” need
fuzzy subsets of U , is a categorical rule. On not be identical with the premise “XisA.”
the other hand, the dispositional entail- Cartesian product. I t should be noted that the generalized
ment principle is an inference rule of the modus ponens is related to the interpola-
form XisA tion rule which was described earlier. An
Yis B additional point that should be noted is
usually ( X is A )
( X , Y) is A x B that the generalized modus ponens may be
A C B
regarded as an instance of the composi-
usually (Xis B ) where (X, Y) is a binary variable and A x tional rule of inference.
In the limiting case where “usually” B is defined by
Dispositional modus ponens. In many
becomes “always,” Equation 14 reduces
applications involving common sense
to Equation 13.
reasoning, the premises in the generalized
In essence, the entailment principle
modus ponens are usuality-qualified. In
asserts that from the proposition “XisA”
such cases, one may employ a dispositional
we can always infer a less specific propo- Projection rule.
version of the modus ponens. I t may be
sition “ X i s B.” For example, from the
expressed as
proposition “Mary is young,” which in its (X, Y) is R
canonical form reads X i s xR usually (X is A )
usually ( Y is B i f X is A )
Age(Mary) is YOUNG where xR, the projection of the binary
usually’ ( Y is B )
relation R on the domain of X, is defined
where YOUNG is interpreted as a fuzzy set by
or, equivalently, as a fuzzy predicate, we where “usually’” is the square of
can infer “Mary is not old,” provided “usually” (see Figure 4). For simplicity,
YOUNG is a subset of the complement of it’s assumed that the premise “ X i s A”
OLD. That is matches the antecedent in the conditional
proposition; also, the conditional propo-
where p R ( u , v )is the membership function sition is interpreted as the statement, “The
April 1988 89
COMPATIBILITY
USUALLY
0 1
Proportion
0 30 50 60
value of the fuzzy conditional probability X i s small In general, the values of a linguistic var-
of B given A is the fuzzy number iable can be generated from a primary
x is small2
USUALLY .” term (for example, “young”) its antonym
(“old”), acollection of modifiers (“not,”
and
Extension principle. The extension prin- “very,” “more or less,” “quite,” “not
ciple plays an important role in fuzzy logic very,” etc.), and the connectives “and”
by providing a mechanism for computing and “or.” For example, one value of
induced constraints. More specifically, “Age” may be “not very young and not
As in the case of the entailment rule, the
assume that a variable X taking values in dispositional version of the extension prin- very old.” Such values can be generated by
a universe of discourse Uis constrained by ciple has the simple form a context-free grammar. Furthermore,
the proposition “ X i s A .” Furthermore, each value of a linguistic variable
assume that f is a mapping from U to V represents a possibility distribution, as
usually ( X is A )
so that X i s mapped intof(X). The ques- shown in Figure 5 for the variable “Age.”
tion is: What is the constraint on f ( X ) usually Cf(X)isf(A)) These possibility distributions may be
which is induced by the constraint on X? computed from the given possibility distri-
The answer provided by the extension The dispositional extension principle butions of the primary term and its anto-
principle may be expressed as the inference plays an important role in inference from nym through the use of attributed
rule common sense knowledge. In particular, grammar techniques.
it is one of the inference rules that play an
XisA essential role in answering the questions An interesting application of the linguis-
posed in the introduction. tic variable is embodied in the fuzzy car
conceived and designed by Sugeno of the
Tokyo Institute of Technology.’ The
where the membership function offlA) is
The linguistic variable car’s fuzzy-logic-based control system lets
it move autonomously along a track with
defined by and its application to rectangular turns and park in a designated
fuzzy control space (see Figure 6). An important feature
is the car’s ability to learn from examples.
A basic concept in fuzzy logic that plays
subject to the condition a key role in many of its applications, espe- The basic idea behind the Sugeno fuzzy
cially in the realm of fuzzy control and car is the following. The controlled vari-
v = f ( u ) , u E U, v E v fuzzy expert systems, is a linguistic able Y , which is the steering angle, is
variable. assumed to be a function of the state vari-
In particular, if the functionfis 1: 1, then A linguistic variable, as its name sug- ables X I , x , , x , , . . . , x,,, which repre-
Equation 15 simplifies to gests, is a variable whose values are words sent the distances of the car from the
or sentences in a natural or synthetic lan- boundaries of the track at a corner (see
guage. For example, “Age” is a linguistic Figure 7). These values are treated as lin-
variable if its values are “young,” “not guistic variables, with the primary terms
where v - ’ is the inverse of v . For young,” “very young,” “old,” “not represented as triangular possibility distri-
example, old,” “very old,” and so on. butions (see Figure 8).
90 COMPUTER
The control policy is represented as a
finite collection of rules of the form kk
R': if (X, is A ' ' ) and . . . (X,is A i ) ,
B
,car
then
Y ' = a b + d l x , + ... +a;x,,
WI Y' + ...+w,Y"
Y = (16)
W' + . . . + w,,
April 1988 91
IF large oil spill or strong acid spill computer employs a parallel architecture. 4. E.H. Mamdani and S. Assilian, “ A Case
THEN emergency is strongly suggested Basically, it has a fuzzy memory, a set of Study on the Application of Fuzzy Set The-
inference engines, a MAX block, and a ory to Automatic Control,” Proc. IFAC
Stochastic ControlSymp., Budapest, 1974.
The fuzziness of such rules is a con- defuzzifier. The computer is designed to
5. Preprints of the Second Congress of the
sequence of the fact that a rule is a sum- process linguistic inputs, for example, International Fuzzy Systems Association,
mary, and summaries, in general, are “more or less small” and “very large,” Tokyo, Japan, 1987.
fuzzy. However, in the context of expert which are represented by analog voltages 6. L.A. Zadeh, “Test-Score Semantics as a
systems and fuzzy logic control, fuzziness on data buses. A binary RAM, an array of Basic for a Computational Approach to the
has the positive effect of reducing the num- registers and a membership function Representation of Meaning,” Literary and
generator form the computer’s fuzzy Linguistic Computing, Vol. 1, 1986, pp.
ber of rules needed to approximately 24-35.
characterize a functional dependence memory.
7. D. Duboisand H . Prade, TheoriedesPos-
between two or more variables. The linguistic inputs are fed to inference sibilites, Masson, Paris, 1985.
engines in parallel, with each rule yielding 8. L.A. Zadeh, “The Role of Fuzzy Logic in
Fuzzy hardware. Several expert system an output. The outputs are aggregated in the Management of Uncertainty in Expert
shells based on fuzzy logic are now com- the MAX block, yielding an overall fuzzy Systems,” F u w Sets and Systems, Vol. 11,
mercially available, among them Reveal output that appears in the output data bus 1983, pp. 199-227.
and Flops.’ The seminal work of Togai as a set of distributed analog voltages. In 9. L.A. Zadeh, “Syllogistic Reasoning in
Fuzzy Logic and its Application to Usual-
and Watanabe at Bell Telephone Labora- intelligent fuzzy control and other appli- ity and Reasoning with Dispositions,”
tories, which resulted in the development cations requiring nonfuzzy commands, the IEEE Trans. Systems, Man, and Cyber-
of a fuzzy logic chip, set the stage for using fuzzy output is fed to a defuzzifier for netics, Vol. SMC-15, 1985, pp. 754-763.
such chips in fuzzy-logic-based expert sys- transformation into crisp output. 10. A. Kaufmann and M. M. Gupta, Introduc-
tems and, more generally, in rule-based Yamakawa’s fuzzy computer may be an tion to Fuzzy Arithmetic, Van Nostrand,
New York, 1985.
systems not requiring a high degree of pre- important step toward a sixth-generation
11. B. Buchanan and E. K. Shortliffe, Rule-
cision.’* More recently, the fuzzy com- computer capable of processing common
Based Expert Systems, Addison-Wesley ,
puter developed by Yamakawa of sense knowledge. This capability is a Reading, Mass., 1984.
Kumamoto University has shown great prerequisite to solving many AI problems 12. M. Togai and H . Watanabe, “Expert Sys-
promise as a general-purpose tool for pro- - for example, handwritten text recogni- tems on a Chip: An Engine for Real-Time
cessing linguistic data at high speed and tion, speech recognition, machine transla- Approximate Reasoning,” IEEE Expert,
with remarkable r o b u s t n e ~ s . ~ tion, summarization, a n d image Vol. 1, 1986, pp. 55-62.
Togai and Watanabe’s fuzzy inference understanding - that do not lend them-
chip consists of four major components: selves to cost-effective solution within the
a rule set memory, an inference processor, bounds of conventional technology. 0
a controller, and 1 / 0 circuitry. In a recent
implementation, a rule set memory is real-
ized by a random-access memory. In the
inference processor, there are 16 data
paths; one data path is laid out for each
rule. All 16 rules on the chip are executed
in parallel. The chip requires 64 clock
Acknowledgment Related Publications
cycles to produce an output. This trans- The research reported in this article has been Adams, E.W., and H.F. Levine, “On the
lates to an execution speed of approxi- supported in part by NASA Grant NCC-2-275 Uncertainties Transmitted from Premises to
and NSF Grant DCR-85 13139. The article was Conclusions in Deductive Inferences,” Syn-
mately 250,000 fuzzy logical inferences per written while the author was a visiting scholar these, Vol. 30, 1975, pp. 429-460.
second (FLIPS) at 16 megahertz clock. A at the Center for the Study of Language and
fuzzy inference accelerator, which is a Information at Stanford University. It is dedi- Baldwin, J.F., and S.Q. Zhou, “A New
coprocessor board for a designated com- cated to the Japanese scientists and engineers Approach to Approximate Reasoning Using a
who have contributed so importantly to the Fuzzy Logic,” Fuzzy Sets and Systems, Vol. 2,
puter, is currently being designed. This
development of fuzzy logic and its applications. 1979, pp. 302-325.
board accommodates the new chips.
In the current implementation, the con- Bellman, R.E., and L.A. Zadeh, “Local and
trol variables are assumed to range over a Fuzzy Logics,” in G. Epstein, ed., Modern Uses
finite set having no more than 31 elements. of Multiple- Valued Logic, Reidel, Dordrecht,
1977, pp. 103-165.
The membership function is quantized at References
16 levels, with 15 representing full mem- 1. L.A. Zadeh, “Fuzzy Algorithms,” Infor- Dubois, D., a n d H . Prade, FuzzySetsandSys-
bership. Once the Togai/Watanabe chip mation and Control, Vol. 12, 1968, pp. tems: Theory and Applications, Academic
becomes available commercially in 1988or 94- 102. Press, New York, 1980.
1989, it should find many uses in both 2. L.A. Zadeh, “ A Rationale for Fuzzy Con-
Goguen, J.A., “The Logic of Inexact Con-
fuzzy-logic-based intelligent controllers trol,” J. Dynamic Systems, Measurement
and Control, Vol. 94, Series G, 1972, pp. cepts,” Synthese, Vol 19, 1969, pp. 325-373.
and expert systems. 3-4.
Yamakawa’s fuzzy computer, whose Klir, G.J., and T.A. Folger, Fuzzy Sets, Uncer-
hardware was built by OMRON Tateise 3. L.A. Zadeh, “Outlineofa New Approach tainty, andlnformation, Prentice Hall, Engle-
to the Analysis of Complex Systems and wood Cliffs, N.J., 1988.
Electronics Corporation, is capable of per- Decision Processes,” IEEE Trans. Systems,
forming fuzzy inference at the very high Man, andcybernetics, Vol. SMC-3,1973, Moisil, G.C., Lectures on the Logic oJFuzzy
speed of 10 megaFLIPS. Yamakawa’s pp. 28-44. Reasoning, Scientific Editions, Bucharest, 1975.
92 COMPUTER
Prade, H., and C . V . Negoita, Fuzzy Logic in
Knowledge Engineering, Verlag TUV Rhein- SOFTWARE ENGINEERS
land, Koln, 1986.
1-
-I DA
Sabatier University, France, in recognition of evaluation
his development of the theory of fuzzy sets. Combined hardware/software
system design methodologies
Zadeh’s address is Computer Science Divi- Transformational approaches to
sion, Dept. of EECS, University of California, automated program generation
Berkeley, C A 94720. Software reusability and reuse
April 1988