Research map / Reinforcement learning / Policy gradient and actor-critic
Non-asymptotic Confidence Intervals of Off-policy Evaluation: Primal and Dual Bounds
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 BONDIC instead of IS-based non-asymptotic bounds | Infinite-horizon off-policy evaluation · Inverted-Pendulum, CartPole, and Type-1 Diabetes simulator | confidence interval width | Helped |
| Use BONDIC instead of bootstrap-based asymptotic bounds | Infinite-horizon off-policy evaluation · Inverted-Pendulum, CartPole, and Type-1 Diabetes simulator | coverage reliability | Helped |
| Use FQE bootstrap instead of BONDIC | Infinite-horizon off-policy evaluation · Experiments with δ = 0.1 | confidence interval width | Helped |
| Use FQE bootstrap instead of BONDIC | Infinite-horizon off-policy evaluation · Experiments with δ = 0.1 | coverage reliability | Hurt |
| Increase the dataset size | BONDIC · Off-policy evaluation benchmarks | confidence interval width | Helped |
| Increase the significance level δ | BONDIC · Off-policy evaluation benchmarks | confidence interval width | Helped |
| Increase the radius r_Q of the Q-function class (ablation) | BONDIC · Synthetic environment | confidence interval width | Mixed |
| Use the dual bound rather than the primal bound | BONDIC · Off-policy evaluation benchmarks | computational efficiency | Helped |
Closest papers by meaning
- Accountable Off-Policy Evaluation With Kernel Bellman Statistics (ICML 2020 · similarity 0.91)
- Simultaneous Statistical Inference for Off-Policy Evaluation in Reinforcement Learning (NeurIPS 2025 · similarity 0.91)
- Off-Policy Interval Estimation with Lipschitz Value Iteration (NeurIPS 2020 · similarity 0.90)
- Statistical Tractability of Off-policy Evaluation of History-dependent Policies in POMDPs (ICLR 2025 · similarity 0.89)
- Model Selection for Off-policy Evaluation: New Algorithms and Experimental Protocol (NeurIPS 2025 · similarity 0.88)
- OPERA: Automatic Offline Policy Evaluation with Re-weighted Aggregates of Multiple Estimators (NeurIPS 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.