A0744
Title: Multilinear common component analysis for tensor data based on Kronecker product approach
Authors: Kohei Yoshikawa - NTT DATA Mathematical Systems Inc. (Japan)
Shuichi Kawano - The University of Electro-Communications (Japan) [presenting]
Abstract: The common component analysis is a multivariate method that extracts a common structure from several datasets. We present multilinear common component analysis (MCCA) based on Kronecker products of mode-wise covariance matrices in order to extract a common structure from multiple tensor datasets. We develop an estimation algorithm and establish the convergence properties of the algorithm. Numerical studies are given to show the effectiveness of MCCA.
