Support Vector Machines (SVMs) versus Multilayer Perception (MLP) in data classification

Abstract

In this paper, we introduce a new kernel function for improving the accuracy of the Support Vector Machines (SVMs) classification. The proposed kernel function is stated in general form and is called Gaussian Radial Basis Polynomials Function (GRPF) that combines both Gaussian Radial Basis Function (RBF) and Polynomial (POLY) kernels. We implement the proposed kernel with a number of parameters associated with the use of the SVM algorithm that can impact the results. A comparative analysis of SVMs versus the Multilayer Perception (MLP) for data classifications is also presented to verify the effectiveness of the proposed kernel function. We seek an answer to the question: “which kernel can achieve a high accuracy classification versus multi-layer neural networks”. The support vector machines are evaluated in comparisons with different kernel functions and multi-layer neural networks by application to a variety of non-separable data sets with several attributes. It is shown that the proposed kernel gives good classification accuracy in nearly all the data sets, especially those of high dimensions. The use of the proposed kernel results in a better, performance than those with existing kernels.