Doubly Robust Off-Policy Value and Gradient Estimation for Deterministic Policies
On the map: Offline RL and imitation learning from data → Offline RL algorithms
Abstract (arXiv)
Offline reinforcement learning, wherein one uses off-policy data logged by a fixed behavior policy to evaluate and learn new policies, is crucial in applications where experimentation is limited such as medicine. We study the estimation of policy value and gradient of a deterministic policy from off-policy data when actions are continuous. Targeting deterministic policies, for which action is a deterministic function of state, is crucial since optimal policies are always deterministic (up to ties). In this setting, standard importance sampling and doubly robust estimators for policy value and gradient fail because the density ratio does not exist. To circumvent this issue, we propose several new doubly robust estimators based on different kernelization approaches. We analyze the asymptotic mean-squared error of each of these under mild rate conditions for nuisance estimators. Specifically, we demonstrate how to obtain a rate that is independent of the horizon length.
Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Placed in Offline RL algorithms, one of 13 directions in Offline RL and imitation learning from data. First posted on arXiv 6 Jun 2020.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2020: 1; 2021: 1; 2022: 2; 2023: 2; 2024: 1; 2025: 2.
9 papers.
- 2020Is Pessimism Provably Efficient for Offline RL?cited by 362
- 2022A Review of Off-Policy Evaluation in Reinforcement Learningcited by 92
- 2024A Causal Framework for Evaluating Deferring Systemscited by 13
- 2025Treatment Effect Estimation for Optimal Decision-Makingcited by 7
- 2023Doubly Robust Proximal Causal Learning for Continuous Treatmentscited by 4
- 2023STEEL: Singularity-aware Reinforcement Learningcited by 3
- 2021Explaining Off-Policy Actor-Critic From A Bias-Variance Perspectivecited by 2
- 2025Breaking Determinism: Stochastic Modeling for Reliable Off-Policy Evaluation in Ad Auctionscited by 2
- 2022Improving Sample Efficiency in Evolutionary RL Using Off-Policy Rankingcited by 0
- 2025Breaking Determinism: Stochastic Modeling for Reliable Off-Policy Evaluation in Ad Auctions3 Dec 2025
- 2025Treatment Effect Estimation for Optimal Decision-Making19 May 2025
- 2024A Causal Framework for Evaluating Deferring Systems29 May 2024
- 2023Doubly Robust Proximal Causal Learning for Continuous Treatments22 Sep 2023
- 2023STEEL: Singularity-aware Reinforcement Learning30 Jan 2023
- 2022A Review of Off-Policy Evaluation in Reinforcement Learning13 Dec 2022
- 2022Improving Sample Efficiency in Evolutionary RL Using Off-Policy Ranking22 Aug 2022
- 2021Explaining Off-Policy Actor-Critic From A Bias-Variance Perspective6 Oct 2021
- 2020Is Pessimism Provably Efficient for Offline RL?30 Dec 2020
Similar, not cited
Close in wording, no citation either way
- 2020Reliable Off-policy Evaluation for Reinforcement Learning
- 2023Double Pessimism is Provably Efficient for Distributionally Robust Offline Reinforcement Learning: Generic Algorithm and Robust Partial Coverage
- 2023On the Sample Complexity of Vanilla Model-Based Offline Reinforcement Learning with Dependent Samples
- 2022RORL: Robust Offline Reinforcement Learning via Conservative Smoothing
- 2024Minimax Optimal and Computationally Efficient Algorithms for Distributionally Robust Offline Reinforcement Learning
- 2026Robust Regularized Policy Iteration under Transition Uncertainty
- 2022Conservative Bayesian Model-Based Value Expansion for Offline Policy Optimization
- 2023Achieving the Asymptotically Optimal Sample Complexity of Offline Reinforcement Learning: A DRO-Based Approach