Targeted Active Learning for Preference-Based Treatment Effects on Multivariate Outcomes

2026-10-02

AI Summary

This study addresses the challenge of eliciting patient preferences in multi-outcome clinical decision-making, where preference acquisition is difficult and annotation costs are prohibitive. To this end, it proposes a goal-oriented active learning framework for target quantity optimization. Unlike conventional approaches that maximize preference information gain, this method models preference rules using Gaussian processes and derives closed-form approximate query criteria, enabling efficient querying by selecting sample pairs that most effectively reduce uncertainty in treatment effects and optimal policies. Experiments on semi-synthetic Parkinson’s disease data demonstrate that, under an equivalent annotation budget, the proposed approach significantly reduces both treatment effect estimation error and policy regret. Ultimately, this work achieves efficient, preference-based estimation of treatment effects.

Abstract

Treatment efficacy is traditionally demonstrated on the basis of a single primary outcome. However, clinical decision-making usually requires consideration of multiple outcomes, balancing expected benefits against potential risks. The relative value assigned to these outcomes varies substantially from one patient to another. Given a preference rule over outcome profiles, treatment effects and optimal policies can be defined and estimated. Such a rule is rarely available in practice: it must itself be estimated from pairwise comparisons of outcome profiles, which are costly to collect from clinical experts. We propose an active learning framework that selects which comparisons to query. Standard criteria maximize the information gained on the preference rule itself. We instead target the quantities of interest, and select the query that most reduces uncertainty on the treatment effect and on the optimal policy induced by the learned rule. Under a Gaussian process model of the preference rule, we derive a closed-form approximation of this criterion. On semi-synthetic data built from a Parkinson's disease cohort with 13 clinical outcomes, our criterion achieves lower treatment effect estimation error and lower policy regret than existing criteria at equal query budget.

Problem

• preference-based treatment effects
• multivariate outcomes
• active learning
• clinical decision-making
• optimal policy

Innovation

• Active Learning
• Preference-Based Treatment Effects
• Gaussian Process
• Optimal Policy
• Multivariate Outcomes