Research map / Bandits and online learning / Multi-armed bandits
MABSplit: Faster Forest Training Using Multi-Armed Bandits
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 adaptive MABSplit node splitting instead of exact brute-force splitting | RF · MNIST | training time (s): 1542.83 $\pm$ 5.837 -> 40.359 $\pm$ 0.246 | Helped |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · MNIST | test accuracy: 0.777 $\pm$ 0.005 -> 0.763 $\pm$ 0.008 | Hurt |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · APS Failure at Scania Trucks | training time (s): 20.542 $\pm$ 0.048 -> 0.455 $\pm$ 0.002 | Helped |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · APS Failure at Scania Trucks | test accuracy: 0.985 $\pm$ 0.0 -> 0.985 $\pm$ 0.0 | No clear effect |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · Forest Covertype | training time (s): 117.351 $\pm$ 0.123 -> 0.88 $\pm$ 0.009 | Helped |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · Forest Covertype | test accuracy: 0.559 $\pm$ 0.028 -> 0.505 $\pm$ 0.004 | Hurt |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · Beijing Multi-Site Air-Quality | training time (s): 138.782 $\pm$ 1.581 -> 67.089 $\pm$ 1.682 | Helped |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · Beijing Multi-Site Air-Quality | test MSE: 1164.576 $\pm$ 0.761 -> 1109.542 $\pm$ 23.776 | Helped |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · SGEMM GPU Kernel Performance | training time (s): 32.606 $\pm$ 0.859 -> 16.51 $\pm$ 0.224 | Helped |
| Use adaptive MABSplit node splitting instead of exact brute-force splitting | RF · SGEMM GPU Kernel Performance | test MSE: 69733.002 $\pm$ 57.401 -> 69493.921 $\pm$ 73.133 | Helped |
| Use adaptive MABSplit node splitting under a fixed computational budget | RF · MNIST | test accuracy: 0.143 $\pm$ 0.026 -> 0.83 $\pm$ 0.002 | Helped |
| Use adaptive MABSplit node splitting under a fixed computational budget | RF · Forest Covertype | test accuracy: 0.514 $\pm$ 0.019 -> 0.675 $\pm$ 0.002 | Helped |
Closest papers by meaning
- DynFrs: An Efficient Framework for Machine Unlearning in Random Forest (ICLR 2025 · similarity 0.87)
- Understanding Randomization in Greedy Model Search (NeurIPS 2026 · similarity 0.87)
- Pushing the Envelope of Gradient Boosting Forests via Globally-Optimized Oblique Trees (CVPR 2022 · similarity 0.87)
- Generalized Random Forests Using Fixed-Point Trees (ICML 2025 · similarity 0.84)
- Minimax Rates and Spectral Distillation for Tree Ensembles (NeurIPS 2026 · similarity 0.84)
- Near-Optimal Decision Trees in a SPLIT Second (ICML 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.