Adapting Neural Networks for the Estimation of Treatment Effects

By Claudia Shi, David M. Blei, Victor Veitch. Neural Information Processing Systems, 2019

503 citations recorded by the source corpus.

This paper addresses the use of neural networks for the estimation of treatment effects from observational data. Generally, estimation proceeds in two stages. First, we fit models for the expected outcome and the probability of treatment (propensity score) for each unit. Second, we plug these fitted models into a downstream estimator of the effect. Neural networks are a natural choice for the models in the first step. The question we address is: how can we adapt the design and training of the neural networks used in the first step in order to improve the quality of the final estimate of the treatment effect? We propose two adaptations based on insights from the statistical literature on the estimation of treatment effects. The first is a new architecture, the Dragonnet, that exploits the sufficiency of the propensity score for estimation adjustment. The second is a regularization procedure, targeted regularization, that induces a bias towards models that have non-parametrically optimal asymptotic properties `out-of-the-box`. Studies on benchmark datasets for causal inference show these adaptations outperform existing methods. Code is available at this http URL.

Subfields: Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/a278c07c8bd2921e59dd862cd91a0540dd340030

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Estimating the Effects of Continuous-valued Interventions using Generative Adversarial Networks (2020, methodology): “In the latter case, methods for handling the selection bias involve propensity weighting (Crump et al., 2008; Alaa et al., 2017; Shi et al., 2019), building sub-populations using tree based methods (Chipman et al.”
Causal machine learning for predicting treatment outcomes (2024, unknown): “This is also called overlap assumption.”
Using Wasserstein Generative Adversarial Networks for the design of Monte Carlo simulations (2019, unknown): “A conventional approach in econometrics is to estimate the distribution fX(·) using kernel density estimation (Silverman [2018], Härdle [1990]).”
Learning Interpretable Models with Causal Guarantees (2019, methodology): “, 2017)— for example, honest trees (Athey and Imbens, 2016), causal forests (Wager and Athey, 2017), propensity score weighting (Austin, 2011; Shi et al., 2019), and causal representations (Johansson et al.”