A NEW INTERPOLATIVE REASONING METHOD IN SPARSE RULE-BASED SYSTEMS

In [7], Yan et al. analyzed Koczy and Hirota's linear interpolative re asoning method presented in [2,3] and found that the reasoning consequ ences by their method sometimes become abnormal fuzzy sets. Thus, they pointed out that a new interpolative reasoning method will be needed which can guarantee that the interpolated conclusion will also be tria ngular-type for a triangular-type observation. In this paper, we exten d the works of [2,3,7] to present a new interpolative reasoning method to deal with fuzzy reasoning in sparse rule-based systems. The propos ed method can overcome the drawback of Koczy and Hirota's method descr ibed in [7]. It can guarantee that the statement ''If fuzzy rules A(1) double right arrow B-1, A(2) double right arrow B-2, and the observat ion A are defined by triangular membership functions, the interpolate d conclusion B will also be triangular-type'' holds. (C) 1998 Elsevie r Science B.V.