Adversarial Debiasing for Unbiased Parameter Recovery
This addresses bias in parameter estimates for researchers using machine-learned predictions in regression, though it is incremental as it adapts existing adversarial methods to a specific problem.
The paper tackles bias in regression coefficient estimates caused by prediction errors from machine learning models used as proxies in social science research, and demonstrates that an adversarial machine learning algorithm can recover true coefficients, as shown in simulations and empirical data on African forest cover.
Advances in machine learning and the increasing availability of high-dimensional data have led to the proliferation of social science research that uses the predictions of machine learning models as proxies for measures of human activity or environmental outcomes. However, prediction errors from machine learning models can lead to bias in the estimates of regression coefficients. In this paper, we show how this bias can arise, propose a test for detecting bias, and demonstrate the use of an adversarial machine learning algorithm in order to de-bias predictions. These methods are applicable to any setting where machine-learned predictions are the dependent variable in a regression. We conduct simulations and empirical exercises using ground truth and satellite data on forest cover in Africa. Using the predictions from a naive machine learning model leads to biased parameter estimates, while the predictions from the adversarial model recover the true coefficients.