A Novel Twin Parametric Support Vector Machine for Large Scale Problem

Dawarwee Makmuang, Rabian Wangkeeree, Cholwich Nattee, Nirattaya Khamsemanan

Abstract

In this paper, we propose a stochastic gradient descent algorithm, called stochastic gradient descent method-based generalized pinball twin parametric support vector machine (SG-TPSVM) to solve data classification problems. This approach is developed by replacing hinge loss function in the conventional twin parametric support vector machine (TPSVM) with a generalized pinball loss function. Moreover, the numerical experiment solved by proposed method is shown.