Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Dynamical Linear Bandits
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 |
|---|---|---|---|
| Account for hidden linear dynamics and persist optimistic actions across epochs | D-LINUCB · Synthetic DLB | cumulative regret | Helped |
| Account for hidden linear dynamics and persist optimistic actions across epochs | D-LINUCB · Real-world advertising simulator | cumulative regret | Helped |
| Ignore the system dynamics and hidden state | LinUCB · Synthetic DLB | cumulative regret | Hurt |
| Ignore the system dynamics and hidden state | D-LINUCB · Synthetic DLB | cumulative regret | Hurt |
| Use lambda = log T instead of lambda = 1 (ablation) | D-LINUCB · Synthetic DLB | cumulative regret | Helped |
| Use lambda = log T instead of lambda = 1 (ablation) | D-LINUCB · Real-world advertising simulator | cumulative regret | Helped |
| Overestimate the spectral-radius bound (ablation) | D-LINUCB · Synthetic DLB | cumulative regret | Hurt |
| Underestimate the spectral-radius bound using rho-bar in {0.1, 0.05} (ablation) | D-LINUCB · Synthetic DLB | cumulative regret | No clear effect |
| Set the spectral-radius bound to zero (ablation) | D-LINUCB · Synthetic DLB | cumulative regret | Hurt |
| Use the human-expert policy that prioritizes instantaneous action effects | Human expert · Real-world advertising simulator | cumulative regret | Hurt |
Closest papers by meaning
- Block Optimism for Nonstationary Bandits with Latent Linear Dynamics (NeurIPS 2026 · similarity 0.90)
- Near-Optimal Stochastic Linear Bandits with Delay (NeurIPS 2026 · similarity 0.89)
- Contextual Linear Bandits with Delay as Payoff (ICML 2025 · similarity 0.88)
- Causal Bandits: The Pareto Optimal Frontier of Adaptivity, a Reduction to Linear Bandits, and Limitations around Unknown Marginals (ICML 2024 · similarity 0.88)
- Testing the Feasibility of Linear Programs with Bandit Feedback (ICML 2024 · similarity 0.87)
- Delay as Payoff in MAB (AAAI 2025 · similarity 0.87)
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.