A new perspective on low-rank optimization
arXiv:2105.05947 · doi:10.1007/s10107-023-01933-9
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 convex relaxations. We invoke the matrix perspective function - the matrix analog of the perspective function - and 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.
arXiv commentMajor revision submitted to Mathematical Programming
References in corpus (9)
- A Matrix Convexity Approach to Some Celebrated Quantum Inequalities
- Semidefinite approximations of the matrix logarithm
- Partial facial reduction: simplified, equivalent SDPs via approximations of the PSD cone
- Regularization vs. Relaxation: A conic optimization perspective of statistical variable selection
- Rank-one Convexification for Sparse Regression
- Mixed-Projection Conic Optimization: A New Paradigm for Modeling Rank Constraints
- Solving Large-Scale Sparse PCA to Certifiable (Near) Optimality
- A unified approach to mixed-integer optimization problems with logical constraints
- 2x2 convexifications for convex quadratic optimization with indicator variables