Low-Rank Principal Eigenmatrix Analysis
Machine Learning
2019-04-30 v1 Machine Learning
Methodology
Abstract
Sparse PCA is a widely used technique for high-dimensional data analysis. In this paper, we propose a new method called low-rank principal eigenmatrix analysis. Different from sparse PCA, the dominant eigenvectors are allowed to be dense but are assumed to have a low-rank structure when matricized appropriately. Such a structure arises naturally in several practical cases: Indeed the top eigenvector of a circulant matrix, when matricized appropriately is a rank-1 matrix. We propose a matricized rank-truncated power method that could be efficiently implemented and establish its computational and statistical properties. Extensive experiments on several synthetic data sets demonstrate the competitive empirical performance of our method.
Keywords
Cite
@article{arxiv.1904.12369,
title = {Low-Rank Principal Eigenmatrix Analysis},
author = {Krishna Balasubramanian and Elynn Y. Chen and Jianqing Fan and Xiang Wu},
journal= {arXiv preprint arXiv:1904.12369},
year = {2019}
}