Regularization path for linear model via net method

  • Authors:

  • Xin-Ze Luan;Yong Liang;Cheng Liu;Zong-Ben Xu;Hai Zhang;Kwong-Sak Leung;Tak-Ming Chan

  • Affiliations:

  • Macau University of Science and Technology, China;Macau University of Science and Technology, China;Macau University of Science and Technology, China;Xi'an Jiaotong University, China;Xi'an Jiaotong University, China;Chinese University of Hong Kong, Hong Kong;Chinese University of Hong Kong, Hong Kong

  • Venue:
  • ICSI'12 Proceedings of the Third international conference on Advances in Swarm Intelligence - Volume Part II
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

We investigate a net regularization method for variable selection in the linear model, which has convex loss function and concave penalty. Meanwhile, the net regularization based on the use of the Lr penalty with $\frac{1}{2}\leq$r ≤1. In the simulation we will demonstrate that the net regularization is more efficient and more accurate for variable selection than Lasso.