Treatment-Response Models for Counterfactual Reasoning with Continuous-time, Continuous-valued Interventions
Abstract
Treatment effects can be estimated from observational data as the difference in potential outcomes. In this paper, we address the challenge of estimating the potential outcome when treatment-dose levels can vary continuously over time. Further, the outcome variable may not be measured at a regular frequency. Our proposed solution represents the treatment response curves using linear time-invariant dynamical systems---this provides a flexible means for modeling response over time to highly variable dose curves. Moreover, for multivariate data, the proposed method: uncovers shared structure in treatment response and the baseline across multiple markers; and, flexibly models challenging correlation structure both across and within signals over time. For this, we build upon the framework of multiple-output Gaussian Processes. On simulated and a challenging clinical dataset, we show significant gains in accuracy over state-of-the-art models.
Cite
@article{arxiv.1704.02038,
title = {Treatment-Response Models for Counterfactual Reasoning with Continuous-time, Continuous-valued Interventions},
author = {Hossein Soleimani and Adarsh Subbaswamy and Suchi Saria},
journal= {arXiv preprint arXiv:1704.02038},
year = {2017}
}
Comments
In Proceedings of the Thirty-Third Conference on Uncertainty in Artificial Intelligence (UAI-2017), Sydney, Australia, August 2017. The first two authors contributed equally to this work