Bayesian preference elicitation for decision support in multiobjective optimization

TL;DR

This paper introduces a Bayesian preference elicitation method for multi-objective optimization that efficiently identifies high-utility solutions by interactively learning decision-maker preferences, reducing the solution set for easier decision-making.

Contribution

It presents a novel Bayesian approach for preference elicitation that adaptively guides the search for optimal solutions in multi-objective problems, with flexible application modes.

Findings

Superior performance in finding high-utility solutions with fewer queries

Effective handling of up to nine objectives in experiments

Provides an open-source implementation for community use

Abstract

We present a novel approach to help decision-makers efficiently identify preferred solutions from the Pareto set of a multi-objective optimization problem. Our method uses a Bayesian model to estimate the decision-maker's utility function based on pairwise comparisons. Aided by this model, a principled elicitation strategy selects queries interactively to balance exploration and exploitation, guiding the discovery of high-utility solutions. The approach is flexible: it can be used interactively or a posteriori after estimating the Pareto front through standard multi-objective optimization techniques. Additionally, at the end of the elicitation phase, it generates a reduced menu of high-quality solutions, simplifying the decision-making process. Through experiments on test problems with up to nine objectives, our method demonstrates superior performance in finding high-utility solutions with a small number of queries. We also provide an open-source implementation of our method to support its adoption by the broader community.