Feature Selection of Support Vector Machine Based on Harmonious Cat Swarm Optimization

Abstract

Cat Swarm Optimization Algorithm (CSO) is an optimization algorithm which proposed in 2006. Indicated by previous studies, CSO has good performance. We proposed a method to improve CSO and presenting a modified CSO named Harmonious-CSO (HCSO). The method is changing the concept of cat alert surroundings in seeking mode of CSO. We change the formula of seeking mode and add a concept of HS algorithm. In this paper, we use Support Vector Machine (SVM) be classifier combine with feature selection to verify the performance of algorithm. For the experimental results, the HCSO algorithm has a better solution than CSO.