Desingularization of bounded-rank matrix sets
From MaRDI portal
Abstract: Conventional ways to solve optimization problems on low-rank matrix sets which appear in great number of applications ignore its underlying structure of an algebraic variety and existence of singular points. This leads to appearance of inverses of singular values in algorithms and since they could be close to it causes certain problems. We tackle this problem by utilizing ideas from the algebraic geometry and show how to desingularize these sets. Our main result is algorithm which uses only bounded functions of singular values and hence does not suffer from the issue described above.
Recommendations
- Matrix optimization over low-rank spectral sets: stationary points and local and global minimizers
- A Riemannian optimization approach to the matrix singular value decomposition
- Controlling singular values with semidefinite programming
- Exact solutions in structured low-rank approximation
- Low-rank optimization on the cone of positive semidefinite matrices
Cites work
- A review of trust region algorithms for optimization
- A Riemannian gradient sampling algorithm for nonsmooth optimization on manifolds
- Dynamical Low‐Rank Approximation
- scientific article; zbMATH DE number 3563286 (Why is no real title available?)
- scientific article; zbMATH DE number 3572315 (Why is no real title available?)
- scientific article; zbMATH DE number 3634395 (Why is no real title available?)
- scientific article; zbMATH DE number 5223994 (Why is no real title available?)
- Low rank approximation. Algorithms, implementation, applications
- Low-rank matrix completion by Riemannian optimization
- Newton’s Method at Singular Points. II
- Numerical solution of saddle point problems
- On a Problem of Weighted Low-Rank Approximation of Matrices
- Preconditioned low-rank Riemannian optimization for linear systems with tensor product structure
- Projection-like retractions on matrix manifolds
- Pymanopt: a Python toolbox for optimization on manifolds using automatic differentiation
- Structured low-rank approximation with missing data
- The Grassmannian variety. Geometric and representation-theoretic aspects
- The Power of Convex Relaxation: Near-Optimal Matrix Completion
- The Rotation of Eigenvectors by a Perturbation. III
Cited in
(16)- A semi-implicit meshless method for incompressible flows in complex geometries
- Learning with tree tensor networks: complexity estimates and model selection
- Mini-workshop: Computational optimization on manifolds. Abstracts from the mini-workshop held November 15--21, 2020 (online meeting)
- Matrix optimization over low-rank spectral sets: stationary points and local and global minimizers
- Analysis of parametric models. Linear methods and approximations
- A multiscale neural network based on hierarchical nested bases
- On the continuity of the tangent cone to the determinantal variety
- Controlling singular values with semidefinite programming
- Jacobi-Davidson method on low-rank matrix manifolds
- Analysis of parametric models for coupled systems
- Finding low-rank solutions via nonconvex matrix factorization, efficiently and provably
- Finding stationary points on bounded-rank matrices: a geometric hurdle and a smooth remedy
- An Apocalypse-Free First-Order Low-Rank Optimization Algorithm with at Most One Rank Reduction Attempt per Iteration
- Group-Invariant Tensor Train Networks for Supervised Learning
- The effect of smooth parametrizations on nonconvex optimization landscapes
- A primal-dual interior point trust region method for second-order stationary points of Riemannian inequality-constrained optimization problems
This page was built for publication: Desingularization of bounded-rank matrix sets
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5373921)
