Surrogate-Assisted Multi-objective Genetic Fuzzy Associative Classification by Multiple Granularity Measures

This paper presents a new surrogate-assisted multi-objective genetic fuzzy associative classification model by learning multiple granularities. The specific method is the hybridization of multi-objective genetic algorithms (MOGAs), radial basis function neural networks (RBFNs), and rough set. We show that our approach requires only a few numbers of fitness evaluations compared to the methods proposed in [34] without compromising to maintain an average classification ability in almost all the datasets considered in this work for evaluation of the model.