Optimization of Ensemble Classifier System based on Multiple Objectives Genetic Algorithm

Abstract

This paper introduces a mechanism to learn optimal combining classifier algorithms associated with features and classifiers from an ensemble system. By using Genetic Algorithm approach that focuses on 3 objectives namely number of correct classified observations and number of selected features and classifiers, optimal solution can be achieved after several interactions of crossover and mutation. We also employ OWA operator in which a weight vector is built by Linear Decreasing (LD) function to average outputs from combining algorithms. Experiments on 2 well-known UCI Machine Learning Repository datasets demonstrate benefits of our approach compared with other state-of-art ensemble method like Decision Template and SCANN as well as all fixed combining algorithms in the ensemble system.