Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Constructive Preference Elicitation Over Hybrid Combinatorial Spaces
Paolo Dragone, Stefano Teso, Andrea Passerini
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 |
|---|---|---|---|
| Use Choice Perceptron instead of SetMargin | Constructive preference elicitation · Synthetic benchmark | regret | Improved |
| Use Choice Perceptron instead of Bayesian competitors | Constructive preference elicitation · Synthetic benchmark | runtime | Improved |
| Use Choice Perceptron instead of SetMargin | PC configuration · PC configuration, uniform users | regret | Improved |
| Use Choice Perceptron instead of SetMargin | PC configuration · PC configuration, normal users | runtime | Improved |
| Increase query set size from k=2 to k=3 (ablation) | Choice Perceptron · Trip planning | average regret | Improved |
| Increase query set size from k=3 to k=4 (ablation) | Choice Perceptron · Trip planning | average regret | Mixed |
| Increase query set size from k=2 to k=3 (ablation) | SetMargin · PC configuration, normal users | regret | Hurt |
| Increase query set size from k=2 to k=3 (ablation) | Choice Perceptron · PC configuration, normal users | regret | Improved |
| Disable hyperparameter tuning (ablation) | SetMargin · Synthetic benchmark | regret | Hurt |
| Increase query set size from k=2 to k=4 (ablation) | Choice Perceptron · Synthetic benchmark | iterations to convergence | No clear effect |
| Increase the time cutoff from 20 seconds to 60 seconds (ablation) | Choice Perceptron · Trip planning | average regret | No clear effect |
Closest papers by meaning
- Optimal Design for Human Preference Elicitation (NeurIPS 2024 · similarity 0.87)
- Preference Learning of Latent Decision Utilities with a Human-like Model of Preferential Choice (NeurIPS 2024 · similarity 0.86)
- Preference Learning with Response Time: Robust Losses and Guarantees (NeurIPS 2025 · similarity 0.86)
- Capturing Individual Human Preferences with Reward Features (NeurIPS 2025 · similarity 0.86)
- Eliciting Human Preferences with Language Models (ICLR 2025 · similarity 0.86)
- Learning from negative feedback, or positive feedback or both (ICLR 2025 · similarity 0.86)
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.
