Effects of heuristic rule generation from multiple patterns in multiobjective fuzzy genetics-Based machine learning

摘要: Fuzzy genetics-based machine learning (FGBML) has frequently been used for fuzzy classifier design. It is one of the promising evolutionary (EML) techniques from viewpoint data mining. This because FGBML can generate accurate classifiers with linguistically interpretable if-then rules. Of course, a tens thousands rules not understandable. Thus, complexity minimization should be considered together accuracy maximization. In previous studies, we proposed hybrid and its multiobjective formulation (MoFGBML) to handle both maximization simultaneously. MoFGBML obtain number non-dominated different tradeoffs between complexity. this paper, focus on heuristic rule generation in improve search performance. original generation, each generated randomly-selected training pattern manner. operation performed at population initialization during evolution. To more generalized according data, propose new where multiple patterns. Through computational experiments using some benchmark sets, discuss effects performance our MoFGBML.