Research map / Probabilistic and scientific modeling / Bayesian optimization
Bayesian Optimization for Categorical and Category-Specific Continuous Inputs
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 Bandit-BO to jointly select a model category and optimize its category-specific hyperparameters | Automated ML across 30 OpenML datasets · 30 OpenML datasets | classification accuracy: 92.25% accuracy; Hyperopt-sklearn (91.48% accuracy), Auto-sklearn (91.82% accuracy), and TPOT (91.77% accuracy) | Helped |
| Use Bandit-BO instead of Hyperopt-sklearn | Automated ML model and hyperparameter selection · wine | classification accuracy: 97.78 (0.01) -> 98.33 (0.00) | Helped |
| Use Bandit-BO instead of the strongest reported baseline | Automated ML model and hyperparameter selection · breast_cancer | classification accuracy: 96.84 (0.00) -> 97.02 (0.01) | Helped |
| Use Bandit-BO instead of Hyperopt-sklearn | Automated ML model and hyperparameter selection · diabetes | classification accuracy: 73.70 (0.02) -> 77.40 (0.01) | Helped |
| Use Bandit-BO instead of Hyperopt-sklearn | Automated ML model and hyperparameter selection · artificial_characters | classification accuracy: 90.94 (0.00) -> 90.47 (0.01) | Hurt |
| Use Bandit-BO instead of Hyperopt-sklearn | Automated ML model and hyperparameter selection · digits | classification accuracy: 98.67 (0.00) -> 98.25 (0.00) | Hurt |
| Use Bandit-BO rather than categorical-input BO baselines | Modified Ackley-5d synthetic function · Modified Ackley-5d, varying batch sizes and category counts | best found function value | Helped |
| Use Bandit-BO rather than categorical-input BO baselines | Feed-forward neural network hyperparameter tuning · Protein structure | negative MSE | Helped |
| Use Bandit-BO rather than competing optimization methods | Synthetic 2d function with one category and one continuous variable · 2d synthetic function | best found function value | Helped |
| Use Bandit-BO rather than competing optimization methods | Modified Alpine-4d synthetic function · Modified Alpine-4d | best found function value | Helped |
Closest papers by meaning
- Bayesian Optimisation over Multiple Continuous and Categorical Inputs (ICML 2020 · similarity 0.92)
- Batched Energy-Entropy acquisition for Bayesian Optimization (NeurIPS 2024 · similarity 0.86)
- Improved Bayes Regret Bounds for Multi-Task Hierarchical Bayesian Bandit Algorithms (NeurIPS 2024 · similarity 0.86)
- Put CASH on Bandits: A Max K-Armed Problem for Automated Machine Learning (NeurIPS 2025 · similarity 0.86)
- Bayesian Optimisation with Unknown Hyperparameters: Regret Bounds Logarithmically Closer to Optimal (NeurIPS 2024 · similarity 0.85)
- Bayesian Optimization from Human Feedback: Near-Optimal Regret Bounds (ICML 2025 · similarity 0.85)
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.