Designing Interpretable Fuzzy Systems

Fuzzy sets and fuzzy logic are powerful tools widely used to represent human knowledge and mimic human reasoning capabilities, being the main constituents of fuzzy systems. Among the different approaches to fuzzy systems, fuzzy rule-based systems represent the one offering a better framework for interpretability considerations. Their applications range from classification to modeling and control. Independently on its purpose, the behavior of a fuzzy rule-based system can be evaluated in terms of its reliability and comprehensibility, two concepts usually represented by accuracy and interpretability in the context of fuzzy systems. Indeed, achieving the highest possible levels of accuracy and interpretability is one of the central aspects of designing fuzzy systems. The present chapter will go through the design process considering its different steps and analyzing the multiple options allowing us to improve interpretability or to achieve a better interpretability-accuracy balance, in the search for interpretable fuzzy systems which represent an interesting approach in the framework of explainable Artificial Intelligence. We will consider questions related to the knowledge extraction and refinement process; some examples are complexity reduction and semantic improvement. We will also analyze other questions linked to the design of the processing structure, such as the effects of applying different aggregation and implication operators.