Comparative Analysis of Support Vector Machine: Employing Various Optimization Algorithms

Support Vector Machines (SVMs) have proved to be good alternative compared to other machine learning techniques specifically for classification problems. The use of optimization methodologies plays a central role in finding solutions of SVMs. In this paper, we present a comparative analysis of study of standard SVM implementing Quadratic Programming (QP) in comparison with SVM employing Evolutionary Algorithm (EvoSVM) and Particle Swam Optimization (PSOSVM) to solve the quadratic optimization problem of SVM. Two class classification problems pertaining to Wisconson Breast Cancer and Bankruptcy prediction using Spanish Banks, Turkish Banks and US Banks data have been analyzed in this study. Based on the datasets used in this study and the results yielded, it is observed that standard SVM outperform other variants of SVM.