Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Anytime Model Selection in Linear Bandits
Parnian Kassraie, Nicolas Emmenegger, Andreas Krause, Aldo Pacchiano
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 ALEXP for online model selection with linearly parametrizable rewards | Linear bandits · Synthetic Legendre-polynomial bandits | cumulative regret | Improved |
| Use ALEXP rather than Explore-then-Commit or Explore-then-Select | Linear bandits with correlated feature maps · Synthetic bandits, M = 55 | cumulative regret | Improved |
| Use ALEXP rather than Corral | Linear bandits · Synthetic bandits, M = 55 and M = 165 | cumulative regret | Improved |
| Increase the number of candidate models | ALEXP · Synthetic bandits, increasing M | cumulative regret | No clear effect |
| Use Lasso reward estimates to emulate full-information feedback in exponential weighting | ALEXP · Linear bandits with M candidate models | regret dependence on M: poly M -> log M | Improved |
| Use a time-uniform Lasso analysis | ALEXP · Linear bandits, arbitrary stopping time | regret guarantee | Improved |
| Use a vanishing exploration probability | ALEXP · Linear bandits, all n ≥ 1 | regret guarantee: \gamma_t = \calO(t^{-1/4}) | Mixed |
Closest papers by meaning
- Sparsity-Agnostic Linear Bandits with Adaptive Adversaries (NeurIPS 2024 · similarity 0.90)
- Causal Bandits: The Pareto Optimal Frontier of Adaptivity, a Reduction to Linear Bandits, and Limitations around Unknown Marginals (ICML 2024 · similarity 0.89)
- Regret minimization in Linear Bandits with offline data via extended D-optimal exploration. (NeurIPS 2026 · similarity 0.88)
- Generalized Linear Bandits: Almost Optimal Regret with One-Pass Update (NeurIPS 2025 · similarity 0.88)
- Robustly Improving Bandit Algorithms with Confounded and Selection Biased Offline Data: A Causal Approach (AAAI 2024 · similarity 0.88)
- Adaptive Regret for Bandits Made Possible: Two Queries Suffice (ICLR 2024 · 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.
