A0396
Title: Estimation of constrained statistical functionals for fair machine learning
Authors: David Benkeser - Emory University (United States) [presenting]
Abstract: Constrained learning has become increasingly important, especially in the realm of algorithmic fairness and machine learning. In these settings, predictive models are developed specifically to satisfy pre-defined notions of fairness. The general problem of constrained statistical machine learning is studied through a statistical functional lens. Learning a function-valued parameter of interest is considered under the constraint that one or several pre-specified real-valued functional parameters equal zero or are otherwise bounded. The constrained functional parameter is characterized as the minimizer of a penalized risk criterion using a Lagrange multiplier formulation. Closed-form solutions for the optimal constrained parameter are often available, providing insight into mechanisms that drive fairness in predictive models. Results also suggest natural estimators of the constrained parameter that can be constructed by combining estimates of unconstrained parameters of the data-generating distribution. Thus, the estimation procedure for constructing fair machine-learning algorithms can be applied in conjunction with any statistical learning approach and off-the-shelf software. The generality of the method is demonstrated by explicitly considering a number of examples of statistical fairness constraints and implementing the approach using several popular learning approaches.
