Research map / Learning theory and algorithms / Kernels and regression theory
Wavelet decompositions of Random Forests - smoothness analysis, sparse approximation and applications
Oren Elisha, Shai Dekel
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 |
|---|---|---|---|
| replace Min-D pruning with wavelet subtree compression | RF 100 trees · UCI Record linkage | nodes required: 123 -> 6 | Improved |
| replace Min-D pruning with wavelet subtree compression | RF 100 trees · UCI CT Slice | nodes required: 77042 -> 5141 | Improved |
| replace Min-D pruning with wavelet subtree compression | RF 100 trees · UCI Titanic | nodes required: 711 -> 34 | Improved |
| replace Min-D pruning with wavelet subtree compression | RF 100 trees · UCI Year Prediction | nodes required: 10657799 -> 9300284 | Improved |
| replace Min-D pruning with wavelet subtree compression | RF 100 trees · UCI Wine quality | nodes required: 39350 -> 29089 | Improved |
| use wavelet approximation instead of standard RF | RF 1000 trees · Spirals original | error: 14.4 -> 12.2 | Improved |
| use wavelet approximation instead of standard RF under 20% label noise | RF 1000 trees · Spirals amplified mislabeling | error: 17.8 -> 13.9 | Improved |
| use wavelet approximation instead of minimal node size pruned RF under label noise | RF 1000 trees · Spirals amplified mislabeling | error: 22.7 -> 13.9 | Improved |
| use wavelet M-term regression instead of Breiman RF | RF 1000 trees · Wine Quality | MSE: 0.4 -> 0.36 | Improved |
| use wavelet regression instead of Denil+S | RF 1000 trees · Wine Quality | MSE: 0.41 -> 0.36 | Improved |
| add wavelet approximation on dataset with bad feature | RF · Poker Hand | regression error | Improved |
| add eight noisy features to the dataset (ablation) | Besov index estimate · indicator of unit sphere | alpha: 0.51 -> 0.51 | No clear effect |
Closest papers by meaning
- Wavelet regression and additive models for irregularly spaced data (NeurIPS 2018 · similarity 0.87)
- The Art of BART: Minimax Optimality over Nonhomogeneous Smoothness in High Dimension (JMLR 2023 · similarity 0.87)
- Scalable Random Wavelet Features: Efficient Non-Stationary Kernel Approximation with Convergence Guarantees (ICLR 2026 · similarity 0.86)
- Minimax Rates and Spectral Distillation for Tree Ensembles (NeurIPS 2026 · similarity 0.83)
- Sampling Binary Data by Denoising through Score Functions (ICML 2025 · similarity 0.83)
- Kernel Regression in Structured Non-IID Settings: Theory and Implications for Denoising Score Learning (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.
