Prediction based on support vector machine for travel choice of high-speed railway passenger in China

Abstract

High-speed railway is a very important part of transportation industry in China, and travel choice has key effect on the development of high-speed railway. Therefore, research on travel behavior of passengers and prediction their travel choice, will offer valuable suggestion for high-speed railway running. In this paper, support vector machine (SVM) is the main method being used to predict. Support vector machine is based on the structural risk minimization principle, and it improves the generalization ability of learning machine to the maximum extent. When solving the limited-sample and nonlinear problems, support vector machine has advantages in predicting. In this research, we get six most important factors, which affect travel choice by the means of questionnaire survey, then use libsvm tool to build prediction model and optimize the train parameters of support vector machine. Finally the prediction accuracy is as high as 91.44%, which shows that support vector machine is good at predicting.