Adversarial Balancing for Causal Inference
This addresses the problem of biased treatment effect estimation in observational studies, offering a more effective solution for researchers and practitioners in fields like healthcare and social sciences, though it is incremental as it builds on existing reweighting techniques.
The paper tackles biases in observational data for causal inference by introducing a novel reweighting method using bi-level optimization and adversarial principles, which achieves improved performance in estimating causal effects compared to previous state-of-the-art methods.
Biases in observational data of treatments pose a major challenge to estimating expected treatment outcomes in different populations. An important technique that accounts for these biases is reweighting samples to minimize the discrepancy between treatment groups. We present a novel reweighting approach that uses bi-level optimization to alternately train a discriminator to minimize classification error, and a balancing weights generator that uses exponentiated gradient descent to maximize this error. This approach borrows principles from generative adversarial networks (GANs) to exploit the power of classifiers for measuring two-sample divergence. We provide theoretical results for conditions in which the estimation error is bounded by two factors: (i) the discrepancy measure induced by the discriminator; and (ii) the weights variability. Experimental results on several benchmarks comparing to previous state-of-the-art reweighting methods demonstrate the effectiveness of this approach in estimating causal effects.