A Unifying Post-Processing Framework for Multi-Objective Learn-to-Defer Problems

TL;DR

This paper introduces a unifying framework for multi-objective learn-to-defer problems, providing a Bayes optimal solution under constraints and demonstrating its effectiveness on real datasets.

Contribution

It develops a generalizable algorithm based on a d-dimensional Neyman-Pearson lemma to optimize learn-to-defer systems with multiple constraints.

Findings

Improved constraint adherence over baselines

Effective estimation of the Bayes optimal solution

Application to COMPAS and ACSIncome datasets

Abstract

Learn-to-Defer is a paradigm that enables learning algorithms to work not in isolation but as a team with human experts. In this paradigm, we permit the system to defer a subset of its tasks to the expert. Although there are currently systems that follow this paradigm and are designed to optimize the accuracy of the final human-AI team, the general methodology for developing such systems under a set of constraints (e.g., algorithmic fairness, expert intervention budget, defer of anomaly, etc.) remains largely unexplored. In this paper, using a $d$-dimensional generalization to the fundamental lemma of Neyman and Pearson (d-GNP), we obtain the Bayes optimal solution for learn-to-defer systems under various constraints. Furthermore, we design a generalizable algorithm to estimate that solution and apply this algorithm to the COMPAS and ACSIncome datasets. Our algorithm shows improvements in terms of constraint violation over a set of baselines.