Identifying Efficient Kernel Function in Multiclass Support Vector Machines

Abstract

Support vector machine (SVM) is a kernel based novel pattern classification method that is significant in many areas like data mining and machine learning. A unique strength is the use of kernel function to map the data into a higher dimensional feature space. In training SVM, kernels and its parameters have very vital role for classification accuracy. Therefore, a suitable kernel design and its parameters should be used for SVM training. In this paper, we present certain kernel functions for multiclass support vector machines and propose the appropriate and optimal kernel for one-versus-one (OAO) and one-versus-all (OAA) multiclass support vector machines. The performance of the one-versus-one and one-versus-all multiclass SVM are illustrated by empirical results and it is evaluated by the parameters like support vectors, support vector percentage, classification error, training error and CPU time. The experimental results demonstrate the ability to use more generalized kernel function and it goes to prove that the polynomial kernel's efficiency in terms of high classification accuracy for several datasets.

Neutron–gamma discrimination based on the support vector machine method

“…here KðX i ; X j Þ is the kernel function [23] used to reflect X i into a higher dimension space. The sequential minimal optimization (SMO) algorithm [24] is used to solve the coefficient α i .…”

“…In this process, the radial basis function (RBF) serves as the kernel function of the SVM due to its excellent nonlinearly reflective ability and simple parameter needed setting [28].…”

Towards a big data framework for analyzing social media content

“…Función Descripción Linear ( , ) = 1 + Kernel lineal. Es una función simple denotada por basada en el parámetro de penalización C (Sangeetha & Kalpana, 2011), dado que ya dicho parámetro controla el intercambio entre la frecuencia de error c y la complejidad de la regla de decisión (Cortes & Vapnik, 1995 (Sangeetha & Kalpana, 2011).…”

Adaptive parameter inversion analysis method of rockfill dam based on harmony search algorithm and mixed multi-output relevance vector machine

“…To improve the computing ability of the M-RVM, the MMRVM incorporating a mixed kernel function is constructed. At present, there are various categories of kernel functions, and the most common kernel functions can be divided into two types (Song et al, 2008;Sangeetha and Kalpana, 2011;Ding et al, 2012): local kernel functions with strong local interpolation ability, such as the Gauss kernel function shown in equation ( 16), and global kernel functions with great generalization ability, such as the polynomial kernel function shown in equation ( 17). To incorporate the advantages of the two Harmony search algorithm types of kernel functions simultaneously, we construct a mixed kernel function, as indicated in equation ( 18).…”