Non-Weighted Evaluation Function in Multi-Objective Problems

Abstract

<jats:p>An evaluation function is proposed to deal with multi-objective problems without weight using a new composition method. Improving the evaluation function by reducing its complexity through discarding the weights. The evaluation function is utilized for the optimization of fuzzy rules.A genetic algorithm is applied as a multi-objective algorithm for fuzzy rules extraction. Simplicity during building fuzzy inference system and reducing the computational complexity is required. The algorithm is applied on AFR data sets.</jats:p>