Research map / Architectures and efficiency / RNNs and normalization
Faster Boosting with Smaller Memory
Julaiti Alafate, Yoav S Freund
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 |
|---|---|---|---|
| weighted sampling (Sparrow) instead of uniform sampling with XGBoost (ablation) | boosted trees · CoverType | test accuracy | Improved |
| Sparrow vs XGBoost (external memory) on 8GB-15GB machines | Sparrow · Splice site | training time: 3x longer | Improved |
| Sparrow vs XGBoost and LightGBM when memory holds the full dataset | Sparrow · Splice site | training time: 1.3 -> 1.1 / 0.5 | Hurt |
| Sparrow vs XGBoost and LightGBM | Sparrow · Bathymetry | training time: 0.4 -> 20.0 / 4.0 | Improved |
| Sparrow vs XGBoost and LightGBM on converged loss | Sparrow · Bathymetry | exponential loss: 0.046 vs 0.054 | Improved |
| Sparrow vs LightGBM on converged loss | Sparrow · Splice site | exponential loss: 0.057 vs 0.053 | Hurt |
| initialize gamma to the maximum advantage of the previous tree's nodes and set a reasonable sample size | Sparrow · large datasets | training time | Improved |
| less memory: 8GB instance vs 15GB instance | Sparrow · Splice site | training time: 8.4 -> 2.9 | Improved |
Closest papers by meaning
- NRGBoost: Energy-Based Generative Boosted Trees (ICLR 2025 · similarity 0.85)
- Optimal Parallelization of Boosting (NeurIPS 2024 · similarity 0.83)
- Revisiting Agnostic Boosting (NeurIPS 2025 · similarity 0.83)
- The Many Faces of Optimal Weak-to-Strong Learning (NeurIPS 2024 · similarity 0.83)
- Sample-Optimal Agnostic Boosting with Unlabeled Data (ICML 2025 · similarity 0.82)
- Analyzing & Reducing the Need for Learning Rate Warmup in GPT Training (NeurIPS 2024 · 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.
