Research map / Learning theory and algorithms / Matrix and tensor methods
A Unified Framework for Structured Low-rank Matrix Learning
Pratik Jawanpuria, Bamdev Mishra
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use RSLM second-order (trust region) solver instead of first-order CG | RSLM · Netflix / ML10m / ML20m matrix completion | test RMSE: 0.8449 -> 0.8443 | No clear effect |
| Use RSLM-tr instead of R3MC | fixed-rank matrix completion · Netflix | test RMSE: 0.8478 -> 0.8443 | Improved |
| Use RSLM-tr instead of nuclear-norm APGL | matrix completion · ML20m | test RMSE: 0.8160 -> 0.7962 | Improved |
| Use RSLM with epsilon-SVR loss instead of l1 loss | robust matrix completion · Netflix | test RMSE: 0.869 -> 0.857 | Improved |
| Replace RMC pseudo-Huber loss with RSLM l1 loss | robust matrix completion · Netflix | test RMSE: 0.868 -> 0.869 | No clear effect |
| Use RSLM instead of BMC | non-negative matrix completion, rank 20 · ML20m | test RMSE: 0.9130 -> 0.8065 | Improved |
| Use RSLM instead of BMC | non-negative matrix completion, rank 10 · ML1m | test RMSE: 0.8529 -> 0.8574 | Hurt |
| Use RSLM instead of BMC on larger data sets | non-negative matrix completion · ML10m / ML20m | test RMSE: 0.8454 -> 0.8142 | Improved |
| Use RSLM instead of BMC for convergence of primal objective | non-negative matrix completion · ML10m, ML20m | objective value vs time | Improved |
| Use RSLM instead of SLRA | Hankel matrix learning · D2, rank 5 | RMSE: 0.0164 -> 0.0059 | Improved |
| Use RSLM instead of SLRA | Hankel matrix learning · D3, rank 5 | RMSE: 0.0149 -> 0.0149 | No clear effect |
| Set rank equal to true system order r0 | RSLM · Hankel D1-D3 | RMSE | Improved |
Closest papers by meaning
- No Spurious Local Minima in Nonconvex Low Rank Problems: A Unified Geometric Analysis (ICML 2017 · similarity 0.88)
- Structured Low-Rank Matrix Factorization: Global Optimality, Algorithms, and Applications (TPAMI 2019 · similarity 0.88)
- Fixed-Rank Supervised Metric Learning on Riemannian Manifold (AAAI 2016 · similarity 0.88)
- A Dual Framework for Low-rank Tensor Completion (NeurIPS 2018 · similarity 0.87)
- Symmetric Matrix Completion with ReLU Sampling (ICML 2024 · similarity 0.87)
- Riemannian coordinate descent algorithms on matrix manifolds (ICML 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
