SOME CONSIDERATIONS ABOUT A LEAST-SQUARES MODEL FOR FUZZY RULES OF INFERENCE

In this paper, the method of least squares is applied to the fuzzy inf erence rules. We begin studying the conditions in which from a fuzzy s et we can build another through the method of least squares. Then we a pply this technique in order to evaluate the conclusions of the genera lized modus ponens and the generalized modus tollens. We present diffe rent theorems and examples that demonstrate the fundamental advantages of the method studied. We conclude presenting a method for obtaining the conclusion when working with fuzzy conditional propositions that c ontain two fuzzy propositions joined by the connective ''and''. (C) 19 98 Elsevier Science B.V. All rights reserved.