Research map / Learning theory and algorithms / Kernels and regression theory
Random-projection ensemble dimension reduction
Wenxing Zhou, Timothy Ivor Cannings
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 |
|---|---|---|---|
| apply the ensemble a second time (RPE2) to the projected data (ablation) | RPE · Model 1a simulation (p=20, n=50) | sin-theta distance: 0.32 -> 0.12 | Improved |
| apply the ensemble a second time (RPE2) (ablation) | RPE · Superconductivity (projection dim 3) | test RMSE: 19.4 -> 17.5 | Improved |
| apply the ensemble a second time (RPE2) (ablation) | RPE · Geographical Origin of Music (projection dim 3) | test RMSE: 16.9 -> 16.1 | Improved |
| apply the ensemble a second time (RPE2) (ablation) | RPE · Residential Building (sales, dim 10) | test RMSE: 124 -> 124 | No clear effect |
| use Cauchy rather than Gaussian projections in sparse models (ablation) | RPE · Model 1a (sparse) | sin-theta distance | Improved |
| use Cauchy rather than Gaussian projections in non-sparse models (ablation) | RPE · Model 1c (no sparsity) | sin-theta distance | Hurt |
| use 50/50 Gaussian-Cauchy mixture projections (ablation) | RPE · Models 1a-1c | sin-theta distance | Improved |
| increase number of groups L (ablation) | RPE · Models 1a, 2, 3 (p=20 and 100, n=200) | sin-theta distance | Improved |
| increase group size M (ablation) | RPE · Models 1-3 (p=20, d=5) | sin-theta distance | Improved |
| use projection dimension d=1 instead of d>d0 (ablation) | RPE · Models 1-3 (p=20) | sin-theta distance | Hurt |
| use MARS as base regression method instead of linear or quadratic least squares (ablation) | RPE · Models 1a and 2 | sin-theta distance | Improved |
| add the ensemble selection step over a single random projection (ablation) | RPE · toy example n=100, p=20 | sin-theta distance | Improved |
Closest papers by meaning
- Enhancing Sufficient Dimension Reduction via Hellinger Correlation (ICML 2024 · similarity 0.85)
- Randomized Dimensionality Reduction for Euclidean Maximization and Diversity Measures (ICML 2025 · similarity 0.84)
- Debiasing Random Oblique Projections for Subsampled OLS and Fast CUR in High Dimensions (NeurIPS 2026 · similarity 0.84)
- Heterogeneous Sufficient Dimension Reduction and Subspace Clustering (ICML 2025 · similarity 0.84)
- High-Dimensional Analysis for Generalized Nonlinear Regression: From Asymptotics to Algorithm (AAAI 2024 · similarity 0.84)
- Dimension-free deterministic equivalents and scaling laws for random feature regression (NeurIPS 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.
