Optimal Prescriptive Trees
Author(s)
Bertsimas, Dimitris J; Dunn, Jack William; Mundru, Nishanth.
DownloadAccepted version (884.8Kb)
Open Access Policy
Open Access Policy
Creative Commons Attribution-Noncommercial-Share Alike
Terms of use
Metadata
Show full item recordAbstract
Motivated by personalized decision making, given observational data {(xᵢ,yᵢ,zᵢ)}ⁿᵢ=1 involving features xᵢ∊ℝᵈ, assigned treatments or prescriptions zᵢ∊{1,...,𝑚}, and outcomes yᵢ∊ℝ, we propose a tree-based algorithm called optimal prescriptive tree (OPT) that uses either constant or linear models in the leaves of the tree to predict the counterfactuals and assign optimal treatments to new samples. We propose an objective function that balances optimality and accuracy. OPTs are interpretable and highly scalable, accommodate multiple treatments, and provide high-quality prescriptions. We report results involving synthetic and real data that show that OPTs either outperform or are comparable with several state-of-the-art methods. Given their combination of interpretability, scalability, generalizability, and performance, OPTs are an attractive alternative for personalized decision making in a variety of areas, such as online advertising and personalized medicine.
Date issued
2019-04Department
Sloan School of Management; Massachusetts Institute of Technology. Operations Research CenterJournal
INFORMS Journal on Optimization
Publisher
Institute for Operations Research and the Management Sciences (INFORMS)
Citation
Bertsimas, Dimitris et al. "Optimal Prescriptive Trees." INFORMS Journal of Optimization 1, 2 (April 2019): 164-183.
Version: Author's final manuscript
ISSN
2575-1484
2575-1492