On Extensions of LARS by Information Geometry : Convex Objectives and τp-Norm
This paper addresses extensions of the Least Angle Regression (LARS) algorithm from two different aspects: (i) from quadratic to more general objectives, and (ii) from `1-norm to `p-norm for p < 1. The equiangular vector, which is the key of LARS, is reproduced in connection with the Riemannian metric induced by the objective function, thereby making the extensions feasible. It is shown, in the case of p < 1, that two types of trajectory ó the c-trajectory and the -trajectory ó need to be distinguished by revealing the discontinuity of the -trajectory.
Masahiro Yukawa Shun-ichi Amari
Dept. Electrical and Electronic Engineering, Niigata University, Japan Brain Science Institute, RIKEN, Japan
国际会议
2011亚太信号与信息处理协会年度峰会(APSIPAASC 2011)
西安
英文
1-6
2011-10-18（万方平台首次上网日期，不代表论文的发表时间）