The paper presents a multi-objective genetic approach to design interpretability-oriented fuzzy rule-based classifiers from data.

The proposed approach allows us to obtain systems with various levels of compromise between their accuracy and interpretability.

Original crossover and mutation operators, as well as chromosome-repairing technique to directly transform the rules are also proposed.

The interpretability measure is based on the arithmetic mean of three components: the average length of rules, the number of active fuzzy sets, and the number of active inputs of the system.

Effectiveness of the proposed technique in various classification problems is confirmed by experimental results.