Research map / Reinforcement learning / MDPs and planning
Off-Policy Evaluation in Partially Observable Environments
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 past and future observations as proxies for hidden states, with invertible conditional probability matrices | POMDP policy evaluation · Synthetic medical environment 1 | estimation bias | Helped |
| Use observed-history importance weights instead of hidden-state importance weights | Importance Sampling · POMDPs without the sufficient condition | estimation error: 0.63α + 0.2γ | Hurt |
| Increase confoundedness | Importance Sampling · Synthetic medical environment 1 | estimation bias | Hurt |
| Apply the Decoupled POMDP estimator where conditional probability matrices are nearly non-invertible | Decoupled POMDP evaluation · Synthetic medical environment 2 | variance | Hurt |
| Use observed-history importance weights | Importance Sampling · Synthetic medical environment 3 | estimation bias | No clear effect |
| Use the Decoupled POMDP estimator instead of the observed-history IS variant | Policy evaluation · Synthetic medical environments 2 and 3 | estimation accuracy | No clear effect |
Closest papers by meaning
- Statistical Tractability of Off-policy Evaluation of History-dependent Policies in POMDPs (ICLR 2025 · similarity 0.94)
- A Minimax Learning Approach to Off-Policy Evaluation in Confounded Partially Observable Markov Decision Processes (ICML 2022 · similarity 0.92)
- Model-based Reinforcement Learning for Confounded POMDPs (ICML 2024 · similarity 0.92)
- Provable Representation with Efficient Planning for Partially Observable Reinforcement Learning (ICML 2024 · similarity 0.90)
- RL in Latent MDPs is Tractable: Online Guarantees via Off-Policy Evaluation (NeurIPS 2024 · similarity 0.89)
- Model Selection for Off-policy Evaluation: New Algorithms and Experimental Protocol (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.