Alternative Strategies in Learning Nonlinear Soft Margin Support Vector Machines

Abstract

The aims of the paper are multifold, to propose a new method to determine a suitable value of the bias corresponding to the soft margin SVM classifier and to experimentally evaluate the quality of the found value against one of the standard expression of the bias computed in terms of the support vectors. Also, it is proposed a variant of the Platt’s SMO algorithm to compute an approximation of the optimal solution of the SVM QP-problem. The new method for computing a more suitable value of the bias is based on genetic search. In order to evaluate the quality of the proposed method from the point of view of recognition and generalization rates, several tests were performed, some of the results being reported in the final section of the paper.

Keywords

• Non-Linear Support Vector Machines
• Kernel Function
• Radial Basis Function
• Soft Margin SVM
• SMO Platt’s Algorithm
• Genetic Search
• Classifier Design and Evaluation