A feature construction approach for genetic iterative rule learning algorithm

摘要: This paper presents a proposal that introduces the use of feature construction in fuzzy rule learning algorithm. is done by means combination two different approaches together with new strategy. The first these consists using relations antecedent rules while second one employs functions rules. Thus, method we propose tries to integrate models so that, strategy allows us start more general and finish process specific ones, are able increase amount information extracted from initial variables. experimental results show proposed obtains good trade-off among accuracy, interpretability time needed get model relation rest algorithms involved comparison.