Robust Bayesian Recourse
This work addresses the need for reliable recourse recommendations in machine learning systems, particularly for users affected by unfavorable decisions, though it appears incremental as it builds on existing recourse methods by adding robustness against model changes.
The paper tackles the problem of generating algorithmic recourse that remains effective despite changes in machine learning model parameters, by introducing a robust Bayesian recourse method that minimizes posterior probability odds ratio and explicitly accounts for data perturbations using a Gaussian mixture ambiguity set. The method demonstrates effectiveness in numerical experiments facing model shifts.
Algorithmic recourse aims to recommend an informative feedback to overturn an unfavorable machine learning decision. We introduce in this paper the Bayesian recourse, a model-agnostic recourse that minimizes the posterior probability odds ratio. Further, we present its min-max robust counterpart with the goal of hedging against future changes in the machine learning model parameters. The robust counterpart explicitly takes into account possible perturbations of the data in a Gaussian mixture ambiguity set prescribed using the optimal transport (Wasserstein) distance. We show that the resulting worst-case objective function can be decomposed into solving a series of two-dimensional optimization subproblems, and the min-max recourse finding problem is thus amenable to a gradient descent algorithm. Contrary to existing methods for generating robust recourses, the robust Bayesian recourse does not require a linear approximation step. The numerical experiment demonstrates the effectiveness of our proposed robust Bayesian recourse facing model shifts. Our code is available at https://github.com/VinAIResearch/robust-bayesian-recourse.