Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Multitask Learning with No Regret: from Improved Confidence Bounds to Active Learning
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 |
|---|---|---|---|
| use improved multitask confidence intervals instead of naive ones in MT-UCB (ablation) | MT-UCB · Synthetic data (d=4, N=5, delta=0.4) online learning | cumulative regret | Helped |
| MT-UCB / AdaMT-UCB multitask learning versus Independent and Single baselines | MT-UCB · Online learning, synthetic and MHC-I | cumulative regret | Helped |
| AdaMT-UCB adapts to unknown task deviation instead of MT-UCB knowing epsilon | MT-UCB · Online learning | cumulative regret | No clear effect |
| MT-AL task selection instead of uniform task sampling and MTS | MT-AL · Active learning, MHC-I drug discovery | cumulative regret | Helped |
| MT-AL versus AE-LSVI | MT-AL · Active learning, MHC-I drug discovery | cumulative regret | No clear effect |
| use improved instead of naive confidence intervals in MT-AL (ablation) | MT-AL · Active learning, MHC-I drug discovery | cumulative regret | Helped |
Closest papers by meaning
- Confidence Weighted Multitask Learning (AAAI 2019 · similarity 0.90)
- Online Multitask Learning with Long-Term Memory (NeurIPS 2020 · similarity 0.90)
- When More Data Doesn't Help: Limits of Adaptation in Multitask Learning (ICML 2026 · similarity 0.89)
- Active Learning from Peers (NeurIPS 2017 · similarity 0.88)
- Fast Rate Bounds for Multi-Task and Meta-Learning with Different Sample Sizes (NeurIPS 2025 · similarity 0.85)
- Bayesian Uncertainty for Gradient Aggregation in Multi-Task Learning (ICML 2024 · similarity 0.84)
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.