Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Optimality of Thompson Sampling with Noninformative Priors for Pareto 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 |
|---|---|---|---|
| Use a prior with k ≥ 2 rather than an optimistic prior | STS · Pareto bandits | expected regret | Helped |
| Use k ≤ 1 rather than k ≥ 2 | STS · Pareto bandits | expected regret | Hurt |
| Use k ≤ 0 | STS · Pareto bandits | expected regret | Hurt |
| Add truncation of the shape estimator, replacing the MLE with min(n, MLE) (ablation) | STS with Jeffreys prior (k = 0) · Pareto bandits | expected regret | Helped |
| Add truncation of the shape estimator, replacing the MLE with min(n, MLE) (ablation) | STS with reference prior (k = 1) · Pareto bandits | expected regret | Helped |
| Add truncation of the shape estimator (ablation) | STS, k = 0 · 4-armed Pareto bandit θ₄ | cumulative regret | Helped |
| Add truncation of the shape estimator (ablation) | STS, k = 1 · 4-armed Pareto bandit θ₄ | cumulative regret | No clear effect |
| Add truncation of the shape estimator (ablation) | STS, k = 3 · 4-armed Pareto bandit θ₄ | cumulative regret | Mixed |
| Use optimistic priors with k < 0 rather than k = 0, 1, or 3 | STS · 4-armed Pareto bandit θ₄ | cumulative regret | Hurt |
| Add the proposed truncation procedure (ablation) | STS with k = -1 or k = -3 · 4-armed Pareto bandit θ₄ | cumulative regret | No clear effect |
| Use k = 3 instead of k = 0 | STS · 4-armed Pareto bandit θ₄ | cumulative regret | Helped |
Closest papers by meaning
- The Choice of Noninformative Priors for Thompson Sampling in Multiparameter Bandit Models (AAAI 2024 · similarity 0.95)
- Thompson Sampling with Less Exploration is Fast and Optimal (ICML 2023 · similarity 0.92)
- Only Pay for What Is Uncertain: Variance-Adaptive Thompson Sampling (ICLR 2024 · similarity 0.90)
- Thompson Sampling For Combinatorial Bandits: Polynomial Regret and Mismatched Sampling Paradox (NeurIPS 2024 · similarity 0.90)
- Thompson Sampling for Multi-Objective Linear Contextual Bandit (NeurIPS 2025 · similarity 0.90)
- Feel-Good Thompson Sampling for Contextual Bandits: a Markov Chain Monte Carlo Showdown (NeurIPS 2025 · similarity 0.89)
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.