A Fast Inference and Type-Reduction Process for Constrained Interval Type-2 Fuzzy Systems

Constrained interval type-2 (CIT2) fuzzy sets have been introduced to preserve interpretability when moving from type-1 (IT2) membership functions. Although they can be used produce systems with enhanced explainability, so far, the latter comes at expense of high computational cost. Specifically, exhaustive type-reduction method for CIT2 Mamdani has shown too slow in practical applications and even current approximation procedure is much slower than modern algorithms IT2 sets. In this article, a novel presented, based on concept switch indices. The algorithm applied real-world classification problem compared other approaches systems. case studies new significantly faster sampling while keeping level operation that characterizes