Research map / Label-efficient and robust learning / Recognition and long-tail
Heuristic Search Algorithm for Dimensionality Reduction Optimally Combining Feature Selection and Feature Extraction
Baokun He, Swair Shah, Crystal Maung, Gordon Arnold, Guihong Wan, Haim Schweitzer
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 |
|---|---|---|---|
| select a feature subset then extract, instead of optimal joint HLR (ablation) | HLR · toy matrix X1 | approximation error: 89.0 -> 77.4 | Improved |
| extract then select, instead of optimal joint HLR (ablation) | HLR · toy matrix X2 | approximation error: 20.44 -> 18.8 | Improved |
| use exact f heuristic instead of f+0.2u (ablation) | HLR search · vehicle Nuclear r1=5 | approximation error: 1402.64 -> 1399.20 | No clear effect |
| use exact f heuristic instead of f+0.8u (ablation) | HLR search · vehicle Nuclear r1=5 | approximation error: 1569.49 -> 1399.20 | Improved |
| use exact f heuristic instead of f+0.2u (ablation) | HLR search · vehicle Spectral r1=5 | approximation error: 326.12 -> 247.58 | Improved |
| add u to f heuristic (f+u) for run time (ablation) | HLR search · vehicle, Schatten 0.25 | run time | Improved |
| exact f heuristic vs exhaustive search (ablation) | HLR search · vehicle, Schatten 0.25 | run time | Improved |
| increase selected features r1 at fixed r=30 (ablation) | HLR · spectf l0 error | l0 error: 0.693 -> 0.647 | Improved |
| increase selected features r1 at fixed r=10 (ablation) | HLR · vehicle Frobenius l0 error | l0 error: 0.562 -> 0.342 | Improved |
| increase extracted features r2 in Greedy HLR (ablation) | Greedy HLR · TechTC01 r1=140 | approximation error: 1205.26 -> 116.84 | Improved |
| run Optimistic Search for more iterations (ablation) | Greedy HLR · TechTC01 5:0 | relative bound: 0.04592 -> 0.03664 | Improved |
| HLR f'=f vs Gu-Eisenstat selection | feature selection · libras Spectral r1=4 | approximation error: 11.863 -> 8.558 | Improved |
Closest papers by meaning
- Unsupervised Feature Selection by Heuristic Search with Provable Bounds on Suboptimality (AAAI 2016 · similarity 0.89)
- Best Subset Selection: Optimal Pursuit for Feature Selection and Elimination (ICML 2025 · similarity 0.84)
- Linear Time Approximation Algorithm for Column Subset Selection with Local Search (NeurIPS 2024 · similarity 0.84)
- Randomized Dimensionality Reduction for Euclidean Maximization and Diversity Measures (ICML 2025 · similarity 0.83)
- Multi-View Unsupervised Column Subset Selection via Combinatorial Search (Student Abstract) (AAAI 2025 · similarity 0.83)
- Optimal Classification Trees for Continuous Feature Data Using Dynamic Programming with Branch-and-Bound (AAAI 2025 · similarity 0.82)
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.
