Off-Policy Evaluation for Missingness-Aware Policies in MDPs with Rewards Missing Not at Random
Ziheng Wei, Annie Qu, Rui Miao
stat.ML, cs.LG
Submitted: 2026-06-18
Comments: Accepted at ICML 2026. 31 pages, 6 figures
Code: https://github.com/NAIVlab/ShadOPE
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- GPT-4 Technical Report
- Generalizing Skills with Semi-Supervised Reinforcement Learning
- Proximal Reinforcement Learning: Efficient Off-Policy Evaluation in Partially Observed Markov Decision Processes
- Off-Policy Reward Shaping with Ensembles
- A Policy Gradient Method for Confounded POMDPs
- Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems
- Evaluating and Learning Optimal Dynamic Treatment Regimes under Truncation by Death
- Deep Reinforcement Learning for Sepsis Treatment
- Empirical Study of Off-Policy Policy Evaluation for Reinforcement Learning
- Off-Policy Evaluation Under Nonignorable Missing Data
- A Fine-grained Analysis of Fitted Q-evaluation: Beyond Parametric Models
- An Introduction to Proximal Causal Learning
- A Review of Off-Policy Evaluation in Reinforcement Learning
- Statistical Tractability of Off-policy Evaluation of History-dependent Policies in POMDPs
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Scalable Posterior Uncertainty for Flexible Density-Based Clustering
- A Transport Geometry of Benign Overfitting
- Efficient quadratic entropy with distance sketches