Research map / Bandits and online learning / Multi-armed bandits
Dueling Bandits with Qualitative Feedback
Liyuan Xu, Junya Honda, Masashi Sugiyama
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 |
|---|---|---|---|
| Estimate pairwise win probabilities directly from qualitative feedback distributions rather than simulate duels with classic DB algorithms | Qualitative dueling bandits · QDB, Condorcet winner | regret | Improved |
| Use Thompson Condorcet sampling instead of classic DB algorithms | QDB · MSLR-WEB10K, Condorcet winner, K=5 | regret | Improved |
| Use Thompson Condorcet sampling instead of classic DB algorithms | QDB · MSLR-WEB10K, Condorcet winner, K=15 | regret | Improved |
| Use Thompson Borda sampling or Borda-UCB instead of DB baselines | QDB · MSLR-WEB10K, Borda winner | regret | Improved |
| Use Thompson Borda sampling rather than Borda-UCB | QDB · Synthetic QDB, K=3, L=4, Borda winner, early rounds | regret | Mixed |
| Use Thompson Borda sampling rather than a method with adequate exploration | QDB · Worst-case Borda instances | regret | Hurt |
| Adopt Borda-UCB's exploration strategy that samples non-most-pulled arms when the UCB arm is not among the most-pulled | Borda-UCB · QDB, Borda winner | regret | Improved |
Closest papers by meaning
- Preference-based Online Learning with Dueling Bandits: A Survey (JMLR 2021 · similarity 0.91)
- Fusing Reward and Dueling Feedback in Stochastic Bandits (ICML 2025 · similarity 0.90)
- Neural Dueling Bandits: Preference-Based Optimization with Human Feedback (ICLR 2025 · similarity 0.88)
- Variance-aware Regret Bounds for Stochastic Contextual Dueling Bandits (ICLR 2024 · similarity 0.88)
- Lipschitz Dueling Bandits over Continuous Action Spaces (NeurIPS 2026 · similarity 0.87)
- Feel-Good Thompson Sampling for Contextual Dueling Bandits (ICML 2024 · similarity 0.87)
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.
