Estimating Counterfactual Distributions under Interference

Shiv Shankar, Ritwik Sinha, Madalina Fiterau
Proceedings of the 4th Machine Learning for Health Symposium, PMLR 259:923-940, 2025.

Abstract

Randomized control trials (RCTs) form a key tool for evaluating medical treatments. However commonly used techniques in RCT based treatment effect estimates suffer from two issues: a) ignoring spillover effects and b) focusing on average (or conditional) average effects. This is partly because evaluating counterfactual distributions is hard; and is further complicated by presence of interference. In this work we propose a new estimator, named RUMI, for estimating distributional quantities like CVaR, QTE from controlled trials under known interference. We provide theoretical justification behind our method and demonstrate its application using synthetic experiments and real data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v259-shankar25a, title = {Estimating Counterfactual Distributions under Interference}, author = {Shankar, Shiv and Sinha, Ritwik and Fiterau, Madalina}, booktitle = {Proceedings of the 4th Machine Learning for Health Symposium}, pages = {923--940}, year = {2025}, editor = {Hegselmann, Stefan and Zhou, Helen and Healey, Elizabeth and Chang, Trenton and Ellington, Caleb and Mhasawade, Vishwali and Tonekaboni, Sana and Argaw, Peniel and Zhang, Haoran}, volume = {259}, series = {Proceedings of Machine Learning Research}, month = {15--16 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v259/main/assets/shankar25a/shankar25a.pdf}, url = {https://proceedings.mlr.press/v259/shankar25a.html}, abstract = {Randomized control trials (RCTs) form a key tool for evaluating medical treatments. However commonly used techniques in RCT based treatment effect estimates suffer from two issues: a) ignoring spillover effects and b) focusing on average (or conditional) average effects. This is partly because evaluating counterfactual distributions is hard; and is further complicated by presence of interference. In this work we propose a new estimator, named RUMI, for estimating distributional quantities like CVaR, QTE from controlled trials under known interference. We provide theoretical justification behind our method and demonstrate its application using synthetic experiments and real data.} }
Endnote
%0 Conference Paper %T Estimating Counterfactual Distributions under Interference %A Shiv Shankar %A Ritwik Sinha %A Madalina Fiterau %B Proceedings of the 4th Machine Learning for Health Symposium %C Proceedings of Machine Learning Research %D 2025 %E Stefan Hegselmann %E Helen Zhou %E Elizabeth Healey %E Trenton Chang %E Caleb Ellington %E Vishwali Mhasawade %E Sana Tonekaboni %E Peniel Argaw %E Haoran Zhang %F pmlr-v259-shankar25a %I PMLR %P 923--940 %U https://proceedings.mlr.press/v259/shankar25a.html %V 259 %X Randomized control trials (RCTs) form a key tool for evaluating medical treatments. However commonly used techniques in RCT based treatment effect estimates suffer from two issues: a) ignoring spillover effects and b) focusing on average (or conditional) average effects. This is partly because evaluating counterfactual distributions is hard; and is further complicated by presence of interference. In this work we propose a new estimator, named RUMI, for estimating distributional quantities like CVaR, QTE from controlled trials under known interference. We provide theoretical justification behind our method and demonstrate its application using synthetic experiments and real data.
APA
Shankar, S., Sinha, R. & Fiterau, M.. (2025). Estimating Counterfactual Distributions under Interference. Proceedings of the 4th Machine Learning for Health Symposium, in Proceedings of Machine Learning Research 259:923-940 Available from https://proceedings.mlr.press/v259/shankar25a.html.

Related Material