Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Multi-Task Learning for Contextual Bandits
Aniket Anand Deshmukh, Urun Dogan, Clay Scott
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 KMTL-UCB with known task similarity instead of Kernel-UCB-Ind, Kernel-UCB-Pool, or KMTL-UCB-Est | Contextual bandits on synthetic news article data · Synthetic news article data | empirical mean regret | Improved |
| Estimate task similarity on the fly with KMTL-UCB-Est instead of using Kernel-UCB-Ind | Contextual bandits on multi-class datasets · Digits, Letter, MNIST, Pendigits, Segment, and USPS | empirical mean regret | Mixed |
| Estimate task similarity on the fly with KMTL-UCB-Est instead of using Kernel-UCB-Ind | Contextual bandits on Digits and Segment · Digits and Segment | empirical mean regret | Improved |
| Estimate task similarity on the fly with KMTL-UCB-Est instead of using Kernel-UCB-Ind | Contextual bandits on MNIST · MNIST | empirical mean regret | Improved |
| Pool all arms into a single estimator instead of modeling task similarity (ablation) | Kernel-UCB on synthetic news article data · Synthetic news article data | empirical mean regret | Hurt |
Closest papers by meaning
- Contextual Bandits with Smooth Regret: Efficient Learning in Continuous Action Spaces (ICML 2022 · similarity 0.89)
- Fast and Sample Efficient Multi-Task Representation Learning in Stochastic Contextual Bandits (ICML 2024 · similarity 0.89)
- Contextual Bandits and Imitation Learning with Preference-Based Active Queries (NeurIPS 2023 · similarity 0.89)
- Active Context Selection Improves Simple Regret in Contextual Bandits (NeurIPS 2026 · similarity 0.89)
- Stochastic Bandits with Context Distributions (NeurIPS 2019 · similarity 0.89)
- Contextual Bandits for Unbounded Context Distributions (ICML 2025 · similarity 0.88)
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.
