Logarithmic Neyman Regret for Adaptive Estimation of the Average Treatment Effect

Ojash Neopane, Aaditya Ramdas, Aarti Singh
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:4303-4311, 2025.

Abstract

Estimation of the Average Treatment Effect (ATE) is a core problem in causal inference with strong connections to Off-Policy Evaluation in Reinforcement Learning. This paper considers the problem of adaptively selecting the treatment allocation probability in order to improve estimation of the ATE. The majority of prior work on adaptive ATE estimation focus on asymptotic guarantees, and in turn overlooks important practical considerations such as the difficulty of learning the optimal treatment allocation as well as hyper-parameter selection. Existing non-asymptotic methods are limited by poor empirical performance and exponential dependence on problem parameters. In order to address these gaps, we propose and analyze the Clipped Second Moment Tracking (ClipSMT) algorithm, a variant of an existing algorithm with strong asymptotic optimality guarantees, and provide finite sample bounds on its Neyman regret. Our analysis shows that, in the superpopulation setting, ClipSMT achieves exponential improvements in Neyman regret on two fronts: improving the dependence on $T$ from $O(\sqrt{T})$ to $O(\log T)$, as well as reducing the exponential dependence on problem parameters to a polynomial dependence—although the setting we consider is slightly less general. We conclude with simulations which show the marked improvement of ClipSMT over existing approaches.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-neopane25a, title = {Logarithmic Neyman Regret for Adaptive Estimation of the Average Treatment Effect}, author = {Neopane, Ojash and Ramdas, Aaditya and Singh, Aarti}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {4303--4311}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/neopane25a/neopane25a.pdf}, url = {https://proceedings.mlr.press/v258/neopane25a.html}, abstract = {Estimation of the Average Treatment Effect (ATE) is a core problem in causal inference with strong connections to Off-Policy Evaluation in Reinforcement Learning. This paper considers the problem of adaptively selecting the treatment allocation probability in order to improve estimation of the ATE. The majority of prior work on adaptive ATE estimation focus on asymptotic guarantees, and in turn overlooks important practical considerations such as the difficulty of learning the optimal treatment allocation as well as hyper-parameter selection. Existing non-asymptotic methods are limited by poor empirical performance and exponential dependence on problem parameters. In order to address these gaps, we propose and analyze the Clipped Second Moment Tracking (ClipSMT) algorithm, a variant of an existing algorithm with strong asymptotic optimality guarantees, and provide finite sample bounds on its Neyman regret. Our analysis shows that, in the superpopulation setting, ClipSMT achieves exponential improvements in Neyman regret on two fronts: improving the dependence on $T$ from $O(\sqrt{T})$ to $O(\log T)$, as well as reducing the exponential dependence on problem parameters to a polynomial dependence—although the setting we consider is slightly less general. We conclude with simulations which show the marked improvement of ClipSMT over existing approaches.} }
Endnote
%0 Conference Paper %T Logarithmic Neyman Regret for Adaptive Estimation of the Average Treatment Effect %A Ojash Neopane %A Aaditya Ramdas %A Aarti Singh %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-neopane25a %I PMLR %P 4303--4311 %U https://proceedings.mlr.press/v258/neopane25a.html %V 258 %X Estimation of the Average Treatment Effect (ATE) is a core problem in causal inference with strong connections to Off-Policy Evaluation in Reinforcement Learning. This paper considers the problem of adaptively selecting the treatment allocation probability in order to improve estimation of the ATE. The majority of prior work on adaptive ATE estimation focus on asymptotic guarantees, and in turn overlooks important practical considerations such as the difficulty of learning the optimal treatment allocation as well as hyper-parameter selection. Existing non-asymptotic methods are limited by poor empirical performance and exponential dependence on problem parameters. In order to address these gaps, we propose and analyze the Clipped Second Moment Tracking (ClipSMT) algorithm, a variant of an existing algorithm with strong asymptotic optimality guarantees, and provide finite sample bounds on its Neyman regret. Our analysis shows that, in the superpopulation setting, ClipSMT achieves exponential improvements in Neyman regret on two fronts: improving the dependence on $T$ from $O(\sqrt{T})$ to $O(\log T)$, as well as reducing the exponential dependence on problem parameters to a polynomial dependence—although the setting we consider is slightly less general. We conclude with simulations which show the marked improvement of ClipSMT over existing approaches.
APA
Neopane, O., Ramdas, A. & Singh, A.. (2025). Logarithmic Neyman Regret for Adaptive Estimation of the Average Treatment Effect. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:4303-4311 Available from https://proceedings.mlr.press/v258/neopane25a.html.

Related Material