4
IRUS TotalDownloads
Altmetric
A new perspective on low-rank optimization
File | Description | Size | Format | |
---|---|---|---|---|
s10107-023-01933-9.pdf | Published version | 839.61 kB | Adobe PDF | View/Open |
Title: | A new perspective on low-rank optimization |
Authors: | Bertsimas, D Cory-Wright, R Pauphilet, J |
Item Type: | Journal Article |
Abstract: | A key question in many low-rank problems throughout optimization, machine learning, and statistics is to characterize the convex hulls of simple low-rank sets and judiciously apply these convex hulls to obtain strong yet computationally tractable relaxations. We invoke the matrix perspective function—the matrix analog of the perspective function—to characterize explicitly the convex hull of epigraphs of simple matrix convex functions under low-rank constraints. Further, we combine the matrix perspective function with orthogonal projection matrices—the matrix analog of binary variables which capture the row-space of a matrix—to develop a matrix perspective reformulation technique that reliably obtains strong relaxations for a variety of low-rank problems, including reduced rank regression, non-negative matrix factorization, and factor analysis. Moreover, we establish that these relaxations can be modeled via semidefinite constraints and thus optimized over tractably. The proposed approach parallels and generalizes the perspective reformulation technique in mixed-integer optimization and leads to new relaxations for a broad class of problems. |
Issue Date: | 1-Nov-2023 |
Date of Acceptance: | 10-Dec-2022 |
URI: | http://hdl.handle.net/10044/1/105265 |
DOI: | 10.1007/s10107-023-01933-9 |
ISSN: | 0025-5610 |
Publisher: | Springer Science and Business Media LLC |
Start Page: | 47 |
End Page: | 92 |
Journal / Book Title: | Mathematical Programming |
Volume: | 202 |
Copyright Statement: | © The Author(s) 2023, corrected publication 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. |
Publication Status: | Published |
Online Publication Date: | 2023-02-18 |
Appears in Collections: | Imperial College Business School |
This item is licensed under a Creative Commons License