Research map / Probabilistic and scientific modeling / Dynamical systems
Sparse Coding and Dictionary Learning With Linear Dynamical Systems
Check 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 sparse coding on the infinite LDS representation instead of the LDS-Martin baseline | LDS classification · Cambridge | accuracy: 88.3 -> 91.7 | Helped |
| Use sparse coding on the infinite LDS representation instead of the LDS-Martin baseline | LDS classification · CK+ | accuracy: 77.3 -> 86.7 | Helped |
| Constrain transition matrices to be symmetric (ablation) | LDS-SC · CK+ | accuracy: 86.7 -> 84.5 | Hurt |
| Constrain transition matrices to be symmetric (ablation) | LDS-SC · Cambridge | accuracy: 91.7 -> 91.7 | No clear effect |
| Add state covariance to symmetric-transition LDS coding (β=0.2) (ablation) | LDSST-SC · UCSD | accuracy: 93.3 -> 90.2 | Hurt |
| Add state covariance to symmetric-transition LDS coding (β=0.6) (ablation) | LDSST-SC · CK+ | accuracy: 84.5 -> 86.7 | Helped |
| Add state covariance to symmetric-transition LDS coding (ablation) | LDSST-SC · Cambridge, UCSD, CK+, SD, SPR, and BDH | accuracy: 85.7, 89.4, 85.4, 98, 95, 99 (β=0.8); 85.7, 89.8, 86.7, 98, 97, 100 (β=0.6); 86.8, 90.2, 86.5, 98, 97, 100 (β=0.2) | Mixed |
| Learn dictionary atoms rather than use randomly selected training systems (ablation) | covLDSST-DL · Cambridge and DynTex++ 9-class subset | accuracy | Helped |
| Increase finite observability order L (ablation) | gLDS-SC · Cambridge, UCSD, and SPR | accuracy | No clear effect |
Closest papers by meaning
- Efficient Optimization for Linear Dynamical Systems with Applications to Clustering and Sparse Coding (NeurIPS 2017 · similarity 0.89)
- Sparse Learning of Dynamical Systems in RKHS: An Operator-Theoretic Approach (ICML 2023 · similarity 0.86)
- A Linearly Convergent Method for Non-Smooth Non-Convex Optimization on the Grassmannian with Applications to Robust Subspace and Dictionary Learning (NeurIPS 2019 · similarity 0.86)
- A Unified Framework for Learning with Nonlinear Model Classes from Arbitrary Linear Samples (ICML 2024 · similarity 0.84)
- Deep-Koopman-KANDy: Dictionary Discovery for Deep-Koopman Operators with Kolmogorov-Arnold Networks for Dynamics (NeurIPS 2026 · similarity 0.84)
- SpectraLDS: Provable Distillation for Linear Dynamical Systems (NeurIPS 2025 · similarity 0.83)
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.