A Unified Framework for Structured Low-Rank Matrix Learning
Abstract
We consider the problem of learning a low-rank matrix, constrained to lie in a linear subspace, and introduce a novel factorization for modeling such matrices. A salient feature of the proposed factorization scheme is it decouples the low-rank and the structural constraints onto separate factors. We formulate the optimization problem on the Riemannian spectrahedron manifold, where the Riemannian framework allows to develop computationally efficient conjugate gradient and trust-region algorithms. Experiments on problems such as standard/robust/non-negative matrix completion, Hankel matrix learning and multi-task learning demonstrate the efficacy of our approach.
Cite
Text
Jawanpuria and Mishra. "A Unified Framework for Structured Low-Rank Matrix Learning." International Conference on Machine Learning, 2018.Markdown
[Jawanpuria and Mishra. "A Unified Framework for Structured Low-Rank Matrix Learning." International Conference on Machine Learning, 2018.](https://mlanthology.org/icml/2018/jawanpuria2018icml-unified/)BibTeX
@inproceedings{jawanpuria2018icml-unified,
title = {{A Unified Framework for Structured Low-Rank Matrix Learning}},
author = {Jawanpuria, Pratik and Mishra, Bamdev},
booktitle = {International Conference on Machine Learning},
year = {2018},
pages = {2254-2263},
volume = {80},
url = {https://mlanthology.org/icml/2018/jawanpuria2018icml-unified/}
}