Research map / Bandits and online learning / Multi-armed bandits
Efficient Ordered Combinatorial Semi-Bandits for Whole-Page Recommendation
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 |
|---|---|---|---|
| add Thompson sampling exploration to network-flow exploitation (alpha 0.5) (ablation) | ordered combinatorial bandit · Yahoo! Webscope, S = 3 | average reward: 0.0846 -> 0.0862 | Helped |
| add Thompson sampling exploration to network-flow exploitation (alpha 0.5) (ablation) | ordered combinatorial bandit · Yahoo! Webscope, S = 5 | average reward: 0.0244 -> 0.0249 | Helped |
| replace epsilon-greedy with Thompson sampling (ablation) | ordered combinatorial bandit · Yahoo! Webscope, S = 3 | average reward: 0.0849 -> 0.0862 | Helped |
| replace epsilon-greedy exploration with Thompson sampling (ablation) | ordered combinatorial bandit · Yahoo! Webscope, S = 5 | average reward: 0.0244 -> 0.0249 | Helped |
| add epsilon-greedy random exploration to exploitation (ablation) | network-flow exploitation · Yahoo! Webscope | average reward: 0.0846 -> 0.0849 (S = 3) | No clear effect |
| shrink posterior variance (alpha 1 -> 0.5) (ablation) | Thompson sampling · Yahoo! Webscope, S = 3 | average reward: 0.0861 -> 0.0862 | No clear effect |
| smaller posterior reshaping alpha (more exploitation) (ablation) | Thompson sampling · synthetic context-free Gaussian process | regret | Mixed |
| model position bias via (action, position) pairs instead of unordered action values (ablation) | epsilon-greedy · synthetic Gaussian process | regret | Helped |
Closest papers by meaning
- Gaussian Process Bandits for Top-k Recommendations (NeurIPS 2024 · similarity 0.88)
- PageRank Bandits for Link Prediction (NeurIPS 2024 · similarity 0.86)
- Bandits Meet Mechanism Design to Combat Clickbait in Online Recommendation (ICLR 2024 · similarity 0.86)
- Optimal and Efficient Contextual Combinatorial Semi-bandits with General Function Approximation (NeurIPS 2026 · similarity 0.86)
- Combinatorial Stochastic-Greedy Bandit (AAAI 2024 · similarity 0.85)
- Neural Combinatorial Clustered Bandits for Recommendation Systems (AAAI 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.