Smoothing Newton Method for L1 Soft Margin Data Classification Problem

  • Authors:
  • Weibing Chen;Hongxia Yin;Yingjie Tian

  • Affiliations:
  • Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences, Beijing, China 100190;Department of Mathematics and Statistics, Minnesota State University Mankato, Mankato, USA MN 56001;Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences, Beijing, China 100190

  • Venue:
  • ICCS 2009 Proceedings of the 9th International Conference on Computational Science
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

A smoothing Newton method is given for solving the dual of the l 1 soft margin data classification problem. A new merit function was given to handle the high-dimension variables caused by data mining problems. Preliminary numerical tests show that the algorithm is very promising.