Preferential Mixture-of-Experts: Interpretable Models that Rely on Human Expertise as much as Possible

TL;DR

This paper introduces Preferential MoE, an interpretable mixture-of-experts model that maximizes human expertise usage in decision-making, with applications in clinical HIV and MDD treatments.

Contribution

It presents a novel model that combines human rules with data-driven classifiers, optimizing interpretability and performance through a coupled multi-objective approach.

Findings

Effective in clinical HIV treatment decision support

Improves interpretability by highlighting when to rely on human rules

Demonstrates utility in managing Major Depressive Disorder

Abstract

We propose Preferential MoE, a novel human-ML mixture-of-experts model that augments human expertise in decision making with a data-based classifier only when necessary for predictive performance. Our model exhibits an interpretable gating function that provides information on when human rules should be followed or avoided. The gating function is maximized for using human-based rules, and classification errors are minimized. We propose solving a coupled multi-objective problem with convex subproblems. We develop approximate algorithms and study their performance and convergence. Finally, we demonstrate the utility of Preferential MoE on two clinical applications for the treatment of Human Immunodeficiency Virus (HIV) and management of Major Depressive Disorder (MDD).