A comparative view of interpolation methods between sparse fuzzy rules

Abstract

Several approaches have been proposed in the last few years for interpolating between sparse fuzzy rules. These proposed methods yield very different results in some cases. This is due to different views on the basic principles underlying the interpolation process. In particular, the problem can be viewed as the one of completing a partially-known mapping associating fuzzy sets with other fuzzy sets, or the one of extending the interpolation mechanism that is applicable to classical functions to fuzzily-specified ones. This paper clarifies the differences between the various methods.