GA-fuzzy modeling and classification: Complexity and performance

The use of genetic algorithms (GAs) and other evolutionary optimization met hods to design fuzzy rules for systems modeling and data classification hav e received much attention in recent literature. Authors have focused on var ious aspects of these randomized techniques, and a whole scale of algorithm s have been proposed. We comment on some recent work and describe a new and efficient two-step approach that leads to good results for function approx imation, dynamic systems modeling and data classification problems. First f uzzy clustering is applied to obtain a compact initial rule-based model. Th en this model is optimized by a real-coded GA subjected to constraints that maintain the semantic properties of the rules. We consider four examples f rom the literature: a synthetic nonlinear dynamic systems model, the iris d ata classification problem, the wine data classification problem, and the d ynamic modeling of a diesel engine turbocharger, The obtained results are c ompared to other recently proposed methods.