A Multi-Task Gaussian Process Model for Inferring Time-Varying Treatment Effects in Panel Data

Yehu Chen, Annamaria Prati, Jacob Montgomery, Roman Garnett
Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, PMLR 206:4068-4088, 2023.

Abstract

We introduce a Bayesian multi-task Gaussian process model for estimating treatment effects from panel data, where an intervention outside the observer’s control influences a subset of the observed units. Our model encodes structured temporal dynamics both within and across the treatment and control groups and incorporates a flexible prior for the evolution of treatment effects over time. These innovations aid in inferring posteriors for dynamic treatment effects that encode our uncertainty about the likely trajectories of units in the absence of treatment. We also discuss the asymptotic properties of the joint posterior over counterfactual outcomes and treatment effects, which exhibits intuitive behavior in the large-sample limit. In experiments on both synthetic and real data, our approach performs no worse than existing methods and significantly better when standard assumptions are violated.

Cite this Paper

BibTeX
@InProceedings{pmlr-v206-chen23d, title = {A Multi-Task Gaussian Process Model for Inferring Time-Varying Treatment Effects in Panel Data}, author = {Chen, Yehu and Prati, Annamaria and Montgomery, Jacob and Garnett, Roman}, booktitle = {Proceedings of The 26th International Conference on Artificial Intelligence and Statistics}, pages = {4068--4088}, year = {2023}, editor = {Ruiz, Francisco and Dy, Jennifer and van de Meent, Jan-Willem}, volume = {206}, series = {Proceedings of Machine Learning Research}, month = {25--27 Apr}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v206/chen23d/chen23d.pdf}, url = {https://proceedings.mlr.press/v206/chen23d.html}, abstract = {We introduce a Bayesian multi-task Gaussian process model for estimating treatment effects from panel data, where an intervention outside the observer’s control influences a subset of the observed units. Our model encodes structured temporal dynamics both within and across the treatment and control groups and incorporates a flexible prior for the evolution of treatment effects over time. These innovations aid in inferring posteriors for dynamic treatment effects that encode our uncertainty about the likely trajectories of units in the absence of treatment. We also discuss the asymptotic properties of the joint posterior over counterfactual outcomes and treatment effects, which exhibits intuitive behavior in the large-sample limit. In experiments on both synthetic and real data, our approach performs no worse than existing methods and significantly better when standard assumptions are violated.} }
Endnote
%0 Conference Paper %T A Multi-Task Gaussian Process Model for Inferring Time-Varying Treatment Effects in Panel Data %A Yehu Chen %A Annamaria Prati %A Jacob Montgomery %A Roman Garnett %B Proceedings of The 26th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2023 %E Francisco Ruiz %E Jennifer Dy %E Jan-Willem van de Meent %F pmlr-v206-chen23d %I PMLR %P 4068--4088 %U https://proceedings.mlr.press/v206/chen23d.html %V 206 %X We introduce a Bayesian multi-task Gaussian process model for estimating treatment effects from panel data, where an intervention outside the observer’s control influences a subset of the observed units. Our model encodes structured temporal dynamics both within and across the treatment and control groups and incorporates a flexible prior for the evolution of treatment effects over time. These innovations aid in inferring posteriors for dynamic treatment effects that encode our uncertainty about the likely trajectories of units in the absence of treatment. We also discuss the asymptotic properties of the joint posterior over counterfactual outcomes and treatment effects, which exhibits intuitive behavior in the large-sample limit. In experiments on both synthetic and real data, our approach performs no worse than existing methods and significantly better when standard assumptions are violated.
APA
Chen, Y., Prati, A., Montgomery, J. & Garnett, R.. (2023). A Multi-Task Gaussian Process Model for Inferring Time-Varying Treatment Effects in Panel Data. Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 206:4068-4088 Available from https://proceedings.mlr.press/v206/chen23d.html.

Related Material