Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Contextual semibandits via supervised learning oracles
Akshay Krishnamurthy, Alekh Agarwal, Miro Dudik
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 |
|---|---|---|---|
| Replace epsilon-Greedy exploration with VCEE variance-constrained exploration (same policy class) | oracle-based contextual semibandit · MSLR and Yahoo! learning-to-rank | average reward | Improved |
| Use a rich policy class (GB5 trees) with VCEE instead of LinUCB linear model | VCEE · MSLR and Yahoo! learning-to-rank | average reward | Improved |
| Use linear policy class with VCEE instead of LinUCB | VCEE · MSLR and Yahoo! learning-to-rank | average reward | Hurt |
| Add semibandit feedback exploitation in EELS (unknown weights) vs ignoring feedback | EELS · theoretical comparison | regret bound exponent in L and B | Improved |
| Replace VCEE marginal-action variance constraint with composite-action variance | VCEE · theoretical | regret bound | Improved |
Closest papers by meaning
- Efficient Contextual Bandits with Uninformed Feedback Graphs (ICML 2024 · similarity 0.86)
- From Contextual Combinatorial Semi-Bandits to Bandit List Classification: Improved Sample Complexity with Sparse Rewards (NeurIPS 2025 · similarity 0.86)
- MultiScale Contextual Bandits for Long Term Objectives (NeurIPS 2025 · similarity 0.86)
- Improved Regret and Contextual Linear Extension for Pandora's Box and Prophet Inequality (NeurIPS 2025 · similarity 0.85)
- Contextual Online Decision Making with Infinite-Dimensional Functional Regression (ICML 2025 · similarity 0.85)
- Neural Dueling Bandits: Preference-Based Optimization with Human Feedback (ICLR 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.
