Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Bayesian Inference of Contextual Bandit Policies via Empirical Likelihood
Jiangrong Ouyang, Mingming Gong, Howard Bondell
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 Bayesian HPD intervals instead of Wilks' chi-squared empirical likelihood intervals | empirical likelihood policy value inference · Simulated contextual bandit, small sample sizes | interval coverage probability | Improved |
| use Bayesian HPD intervals instead of Wilks' intervals | baseline policy value inference · Simulated contextual bandit, baseline policy, small sample sizes | coverage overcoverage | Improved |
| use Bayesian HPD intervals instead of Wilks' intervals | empirical likelihood policy value inference · Simulated contextual bandit, small sample sizes | interval width: Wilks' intervals approximately 10% wider than HPD | Improved |
| increase sample size | Wilks' and HPD intervals · Simulated contextual bandit | coverage probability | Improved |
| use joint posterior (policy value vector) rather than independent per-policy inference | policy comparison · Simulated contextual bandit, n=100 | policy comparison | Improved |
| use univariate posterior for the difference d instead of the joint posterior | policy comparison · Adolescent BMI data | posterior probability of improvement: 0.92 vs 0.92 | No clear effect |
Closest papers by meaning
- Empirical Likelihood for Contextual Bandits (NeurIPS 2020 · similarity 0.92)
- Logarithmic Smoothing for Pessimistic Off-Policy Evaluation, Selection and Learning (NeurIPS 2024 · similarity 0.88)
- $\mathtt{VITS}$ : Variational Inference Thompson Sampling for contextual bandits (ICML 2024 · similarity 0.88)
- Contextual Thompson Sampling via Generation of Missing Data (NeurIPS 2025 · similarity 0.88)
- Diffusion Models Meet Contextual Bandits (NeurIPS 2025 · similarity 0.88)
- The Choice of Noninformative Priors for Thompson Sampling in Multiparameter Bandit Models (AAAI 2024 · 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.
