Deconfounding Temporal Autoencoder: Estimating Treatment Effects over Time Using Noisy Proxies
This addresses the challenge of unbiased treatment effect estimation in fields like medicine where true confounders are often unobserved, representing an incremental advance by combining existing techniques with causal regularization.
The paper tackles the problem of estimating individualized treatment effects over time from observational data with noisy proxy measurements instead of true confounders, and demonstrates that their Deconfounding Temporal Autoencoder method improves over state-of-the-art benchmarks by a substantial margin.
Estimating individualized treatment effects (ITEs) from observational data is crucial for decision-making. In order to obtain unbiased ITE estimates, a common assumption is that all confounders are observed. However, in practice, it is unlikely that we observe these confounders directly. Instead, we often observe noisy measurements of true confounders, which can serve as valid proxies. In this paper, we address the problem of estimating ITE in the longitudinal setting where we observe noisy proxies instead of true confounders. To this end, we develop the Deconfounding Temporal Autoencoder, a novel method that leverages observed noisy proxies to learn a hidden embedding that reflects the true hidden confounders. In particular, the DTA combines a long short-term memory autoencoder with a causal regularization penalty that renders the potential outcomes and treatment assignment conditionally independent given the learned hidden embedding. Once the hidden embedding is learned via DTA, state-of-the-art outcome models can be used to control for it and obtain unbiased estimates of ITE. Using synthetic and real-world medical data, we demonstrate the effectiveness of our DTA by improving over state-of-the-art benchmarks by a substantial margin.