Soft Computing and Its Applications, Volume One

470 Soft Computing and Its Applications

where for i=1, 2, …, n and for j = 1, 2, …, m. Then the gen-

eral inference in the absence of any input is given by:

11 1 2 2 2 m m m

12 n 1 1 2 n 2 1 2 n m

g{f[ (A ,A ,[rA ),B ],f[ (A , A ,f[A ), B ],.....f[ (A,A,f[A,B]}µµ µ

(4.52)

where is any antecedent combiner, f is any function representing the rule firing and

g is the aggregation.

RULES WITH THE SAME CONSEQUENTS

In many cases, the number of fuzzy sets (membership functions) defined on the single

output domain, say r, is typically much less than the number of rules m, that is, r

m. To eliminate this redundancy, we propose a new type of rule ...