Policy Learning for Many Outcomes of Interest: Combining Optimal Policy Trees with Multi-objective Bayesian Optimisation
Computational Economics · 2025-08 · 66 (2) · 971-1001
DOI: 10.1007/s10614-024-10722-1 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computational Economics, online erschienen am 03.10.2024. Verfasst von Patrick Rehill und 1 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Australian National University.
Abstract (Originalsprache)
Methods for learning optimal policies use causal machine learning models to create human-interpretable rules for making choices around the allocation of different policy interventions. However, in realistic policy-making contexts, decision-makers often care about trade-offs between outcomes, not just single-mindedly maximising utility for one outcome. This paper proposes an approach termed Multi-Objective Policy Learning (MOPoL) which combines optimal decision trees for policy learning with a multi-objective Bayesian optimisation approach to explore the trade-off between multiple outcomes. It does this by building a Pareto frontier of non-dominated models for different hyperparameter settings which govern outcome weighting. The method is applied to a real-world case-study of pricing targetting subsididies for anti-malarial medication in Kenya.
Förderung
- Australian National University
Zitieren
APA
Rehill, P., & Biddle, N. (2025). Policy Learning for Many Outcomes of Interest: Combining Optimal Policy Trees with Multi-objective Bayesian Optimisation. Computational Economics, 66(2), 971-1001. https://doi.org/10.1007/s10614-024-10722-1
Vancouver
Rehill P, Biddle N. Policy Learning for Many Outcomes of Interest: Combining Optimal Policy Trees with Multi-objective Bayesian Optimisation. Computational Economics. 2025;66(2):971-1001. doi:10.1007/s10614-024-10722-1
BibTeX
@article{Rehill2025,
author = {Rehill, Patrick and Biddle, Nicholas},
title = {Policy Learning for Many Outcomes of Interest: Combining Optimal Policy Trees with Multi-objective Bayesian Optimisation},
journal = {Computational Economics},
year = {2025},
volume = {66},
number = {2},
pages = {971--1001},
doi = {10.1007/s10614-024-10722-1}
}
EUVIMED