Differentially private learners for heterogeneous treatment effects
Keywords: Causality, differential privacy, treatment effect estimation
Abstract: Patient data is widely used to estimate heterogeneous treatment effects and understand the effectiveness and safety of drugs. Yet, patient data includes highly
sensitive information that must be kept private. In this work, we aim to estimate
the conditional average treatment effect (CATE) from observational data under
differential privacy. Specifically, we present DP-CATE, a novel framework for
CATE estimation that is *doubly robust* and ensures *differential privacy* of the estimates. For this, we build upon non-trivial tools from semi-parametric and robust statistics to exploit the connection between privacy and model robustness.
Our framework is highly general and applies to any two-stage CATE meta-learner
with a Neyman-orthogonal loss function. It can be used with all machine learning models employed for nuisance estimation. We further provide an extension
of DP-CATE where we employ RKHS regression to release the complete doubly
robust CATE function while ensuring differential privacy. We demonstrate the effectiveness of DP-CATE across various experiments using synthetic and real-world
datasets. To the best of our knowledge, we are the first to provide a framework for
CATE estimation that is doubly robust and differentially private.
Primary Area: causal reasoning
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 3221
Loading