Evolutionary algorithms have been successfully used in many studies to design accurate and interpretable fuzzy systems under the name of genetic fuzzy systems. Recently evolutionary multiobjective algorithms have been used for interpretability-accuracy tradeoff analysis of fuzzy systems. We first review a wide range of related studies to multiobjective genetic fuzzy systems. Then we illustrate multiobjective design of fuzzy systems through computational experiments on some benchmark data sets. Finally we point out promising future research directions.
Published in:
Fuzzy Systems Conference, 2007. FUZZ-IEEE 2007. IEEE International
Date of Conference: 23-26 July 2007