Affordable Access

Fifth-order canonical polyadic decomposition with partial symmetry via joint diagonalization for combined independent component analysis and canonical / Parallel factor analysis

Authors
  • Gong, Xiao-Feng
  • Wang, Cheng-Yuan
  • Hao, Ya-Na
  • Lin, Qiu-Hua
Type
Preprint
Publication Date
Nov 05, 2013
Submission Date
Nov 05, 2013
Identifiers
arXiv ID: 1311.1040
Source
arXiv
License
Yellow
External links

Abstract

Recently, there has been a trend to combine independent component analysis and canonical / parallel factor analysis (ICA-CPA) for an enhanced robustness for the computation of CPA, and ICA-CPA could be further converted into the problem of canonical polyadic decomposition (CPD) of a 5th-order partially symmetric tensor, by calculating the 4th-order cumulant of a trilinear mixture. In this study, we propose a new 5th-order CPD algorithm constrained with partial symmetry using joint diagonalization. As the main steps involved in the proposed algorithm undergo no updating iterations for the loading matrices, it is much faster than the existing algorithm based on alternating least squares and enhanced line search, and therefore could be used as a nice initialization for the latter. Simulation results are given to examine the performance of the proposed algorithm.

Report this publication

Statistics

Seen <100 times