Research map / Learning theory and algorithms / Matrix and tensor methods
Linear Convergence of Gradient Methods for Estimating Structured Transition Matrices in High-dimensional Vector Autoregressive Models
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 PGD instead of FNSL | Sparse transition matrix estimation · Sparse VAR, d=100, n=1000 | TPR (%): 73.64 -> 79.49 | Helped |
| Use PGD instead of FNSL | Sparse transition matrix estimation · Sparse VAR, d=100, n=2000 | total time (s): 183.79 -> 6.14 | Helped |
| Use PGD instead of FNSL | Sparse transition matrix estimation · Sparse VAR, d=100, n=1500 | estimation error (EE): 0.417 -> 0.396 | Helped |
| Increase sample size from 1000 to 2000 | PGD · Sparse VAR, d=100 | TPR (%): 79.49 -> 85.82 | Helped |
| Use AltPGD instead of FNSL | Sparse plus low-rank transition matrix estimation · d=100, n=1500 | total time (s): 309.76 -> 19.16 | Helped |
| Use AltPGD instead of FNSL | Sparse plus low-rank transition matrix estimation · d=100, n=2500 | TPR (%): 77.49 -> 83.19 | Helped |
| Use AltPGD instead of FNSL | Sparse plus low-rank transition matrix estimation · d=100, n=2000 | TPR (%): 74.65 -> 81.06 | Helped |
| Use PGD instead of FNSL | Low-rank transition matrix estimation · d=100, n=8000, r=2 | convergence rate | Helped |
| Use AltPGD instead of FNSL | Sparse plus low-rank estimation · d=100, n=8000 | convergence rate | Helped |
| Increase sample size and dimension | PGD and AltPGD · Low-rank and superposition settings | squared estimation error | Helped |
Closest papers by meaning
- Statistical Guarantees for High-Dimensional Stochastic Gradient Descent (NeurIPS 2025 · similarity 0.87)
- Estimating Structured Vector Autoregressive Models (ICML 2016 · similarity 0.87)
- Provable Acceleration of Nesterov's Accelerated Gradient for Asymmetric Matrix Factorization and Linear Neural Networks (NeurIPS 2024 · similarity 0.85)
- The High Line: Exact Risk and Learning Rate Curves of Stochastic Adaptive Learning Rate Algorithms (NeurIPS 2024 · similarity 0.85)
- High-dimensional Limit of SGD for Diagonal Linear Networks (NeurIPS 2026 · similarity 0.85)
- Block Acceleration Without Momentum: On Optimal Stepsizes of Block Gradient Descent for Least-Squares (ICML 2024 · similarity 0.84)
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.