Improving interpretability in approximative fuzzy models via multi-objective evolutionary algorithms

Current research lines in fuzzy modeling mostly tackle with improving the accuracy in descriptive models, and the improving of the interpretability in approximative models. This paper deals with the second issue approaching the problem by means of multi-objective optimization in which accurate and interpretability criteria are simultaneously considered. Evolutionary Algorithms are specially appropriated for multi-objective optimization because they can capture multiple Pareto solutions in a single run of the algorithm. We propose a multi-objective evolutionary algorithm to find multiple Pareto solutions (fuzzy models) showing a trade-off between accuracy and interpretability. Additionally neural network based techniques in combination with ad hoc techniques for interpretability improving, are incorporated into the multi-objective evolutionary algorithm in order to improve the efficacy of the algorithm.