Research map / Probabilistic and scientific modeling / Bayesian optimization
Efficient Automatic CASH via Rising 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 |
|---|---|---|---|
| alternate MAB-based algorithm selection with per-algorithm BO instead of one joint BO over all algorithms | SMAC on CASH · PC4 (K=16, 500 trials) | validation accuracy: 93.63 -> 95.02 | Helped |
| eliminate poor algorithms using Rising Bandit bounds | uniform allocation (AVG) · PC4 K=16 | validation accuracy: 93.39 -> 95.02 | Helped |
| use Rising Bandits for algorithm selection | SMAC · 10 OpenML datasets | trials to same accuracy: 12.6x average speedup | Helped |
| use Rising Bandits vs adaptive BOHB | BOHB · 10 OpenML datasets | speedup: 2.4x to 15.7x | Helped |
| use non-stationary rising bandit with best-reward objective instead of stationary bandits | UCB/CMAB/Softmax · 30 OpenML datasets | validation accuracy: best on 26 of 30 datasets | Helped |
| rising bandits on test set | SMAC/TPE/bandit baselines · 30 OpenML datasets | test accuracy: best on 20 of 30 datasets | Mixed |
| cost-aware Rising Bandits in AutoML system | Auto-Sklearn/HPSK/TPOT · 12 OpenML datasets, 2h | test accuracy: 72.33 -> 82.60 on AMAZON | Helped |
Closest papers by meaning
- Put CASH on Bandits: A Max K-Armed Problem for Automated Machine Learning (NeurIPS 2025 · similarity 0.91)
- Interactive Hyperparameter Optimization in Multi-Objective Problems via Preference Learning (AAAI 2024 · similarity 0.84)
- Adaptive Regret for Bandits Made Possible: Two Queries Suffice (ICLR 2024 · similarity 0.83)
- Beyond Primal-Dual Methods in Bandits with Stochastic and Adversarial Constraints (NeurIPS 2024 · similarity 0.82)
- High-dimensional Linear Bandits with Knapsacks (ICML 2024 · similarity 0.82)
- Bayesian Optimisation with Unknown Hyperparameters: Regret Bounds Logarithmically Closer to Optimal (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.