A Fast and Accurate Rule-Base Generation Method for Mamdani Fuzzy Systems

Liviu-Cristian Dutu (1) , Gilles Mauris (1) , Philippe Bolon (1)

1 LISTIC - Laboratoire d'Informatique, Systèmes, Traitement de l'Information et de la Connaissance

The problem of learning fuzzy rule-bases is analyzed from the perspective of finding a favorable balance between the accuracy of the system, the speed required to learn the rules, and finally, the interpretability of the rule-bases obtained. Therefore, we introduce a complete design procedure to learn and then optimize the system rule-base, called the Precise and Fast Fuzzy Modeling approach. Under this paradigm, fuzzy rules are generated from numerical data using a parameterizable greedy-based learning method called Selection-Reduction, whose accuracy-speed efficiency is confirmed through empirical results and comparisons with reference methods. Qualitative justification for this method is provided based on the co-action between fuzzy logic and the intrinsic properties of greedy algorithms. To complete the Precise and Fast Fuzzy Modeling strategy, we finally present a rule-base optimization technique driven by a novel rule redundancy index which takes into account the concepts of distance between rules and the influence of a rule over the dataset. Experimental results show that the proposed index can be used to obtain compact rule-bases which remain very accurate, thus increasing system interpretability.

Mots clés: fuzzy modeling; inductive rule learning; double-consequent linguistic rules; accuracy-speed trade; greedy rule selection; interpretability–accuracy trade-off; rule-base reduction

Domaines: Traitement du signal et de l'image [eess.SP]

Liviu-Cristian Dutu, Gilles Mauris, Philippe Bolon. A Fast and Accurate Rule-Base Generation Method for Mamdani Fuzzy Systems. IEEE Transactions on Fuzzy Systems, 2018, pp.715-733. ⟨10.1109/TFUZZ.2017.2688349⟩. ⟨hal-01756513⟩