Patrick Rehill, Nicholas Biddle
arXiv 13 Dec 2022 · Machine Learning · publishedComputational Economics (2024) · 1 citations (OpenAlex)
arXiv:2212.06312 · PDF · DOI · OpenAlex · Extracted main text
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 key here is that a low-cost greedy tree can be an accurate proxy for the very computationally costly optimal tree for the purposes of making decisions which means models can be repeatedly fit to learn a Pareto frontier. The method is applied to a real-world case-study of non-price rationing of anti-malarial medication in Kenya.
appendix boundary found by appendix_titled_section at “Appendix A - Variables used in application” · 97% of the source is main text. Read the extracted text to check this.
The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.
| Reference | Intensity | Mentions | Sections | Main text | |
|---|---|---|---|---|---|
| 1 | athey_policy_2020 APACrefauthors Athey, S. \ Wager, S. APACrefauthor… 202009 | 1.000 | 9 | 4 | 100% |
| 2 | sverdrup_policytree_2020 APACrefauthors Sverdrup, E. , Kanodia, A. ,… (2020) 202006 | 1.000 | 5 | 3 | 100% |
| 3 | zhou_offline_2023 APACrefauthors Zhou, Z. , Athey, S. \ Wager, S. AP… (2022) 202301 | 1.000 | 5 | 3 | 100% |
| 4 | athey_generalized_2019 APACrefauthors Athey, S. , Tibshirani, J. \ W… (2019) 201904 | 0.928 | 4 | 3 | 100% |
| 5 | daulton_parallel_2021 APACrefauthors Daulton, S. , Balandat, M. \ Ba… (2021) 202110 | 0.811 | 4 | 2 | 100% |
| 6 | bertsimas_optimal_2017 APACrefauthors Bertsimas, D. \ Dunn, J. APACr… (2017) 201707 | 0.644 | 2 | 2 | 100% |
| 7 | bertsimas_optimal_2019 APACrefauthors Bertsimas, D. , Dunn, J. \ Mun… (2018) 201904 | 0.644 | 2 | 2 | 100% |
| 8 | chernozhukov_doubledebiased_2018 APACrefauthors Chernozhukov, V. , C… (2018) 201802 | 0.644 | 2 | 2 | 100% |
| 9 | kunzel_metalearners_2019 APACrefauthors Künzel, S R. , Sekhon, J S.… (2019) 201903 | 0.644 | 2 | 2 | 100% |
| 10 | morales-hernandez_survey_2022 APACrefauthors Morales-Hernández, A. ,… (2022) 202211 | 0.644 | 2 | 2 | 100% |
Showing the top 10 of 52 scored citations.