Title :
Prediction of Railway Passenger Traffic Volume Based on Weighted LS-SVM
Author :
Han, Hu ; Dang, Jian-wu ; Ren, En-En
Author_Institution :
Sch. of Math., Phys. & Software Eng., Lanzhou Jiaotong Univ., Lanzhou
Abstract :
In prediction of railway passenger traffic volume based on support vector regression, different input points make different contribution to the predictive function. A new prediction method for railway passenger volume, named weighted LS-SVM, is presented in this paper, different weighting factors are assigned to each input points by the linear interpolation function. The railway passenger volume from 1985 to 2002 are used and the results show that the weighted LS-SVM outperforms the standard LS-SVM.
Keywords :
forecasting theory; interpolation; least squares approximations; railways; support vector machines; least squares approximations; linear interpolation function; railway passenger traffic volume prediction; support vector regression; Automation; Interpolation; Least squares methods; Mathematics; Physics computing; Prediction methods; Rail transportation; Software engineering; Support vector machines; Traffic control;
Conference_Titel :
Intelligent Computation Technology and Automation (ICICTA), 2008 International Conference on
Print_ISBN :
978-0-7695-3357-5
DOI :
10.1109/ICICTA.2008.17
