Optimal policy trees

被引：0
作者：
Maxime Amram
Jack Dunn
Ying Daisy Zhuo
机构：
[1] Interpretable AI,
来源：
Machine Learning | 2022年 / 111卷
关键词：
Machine learning; Decision trees; Prescriptive decision making;
D O I：
暂无
中图分类号：
学科分类号：
摘要：
We propose an approach for learning optimal tree-based prescription policies directly from data, combining methods for counterfactual estimation from the causal inference literature with recent advances in training globally-optimal decision trees. The resulting method, Optimal Policy Trees, yields interpretable prescription policies, is highly scalable, and handles both discrete and continuous treatments. We conduct extensive experiments on both synthetic and real-world datasets and demonstrate that these trees offer best-in-class performance across a wide variety of problems.
引用
收藏
页码：2741 / 2768
页数：27
相关论文
共 32 条
[1]  
Aglin G(2020)Learning optimal decision trees using caching branch-and-bound search Proceedings of the AAAI Conference on Artificial Intelligence 34 3146-3153
[2]  
Nijssen S(2017)Optimal classification trees Machine Learning 106 1039-1082
[3]  
Schaus P(2019)Optimal prescriptive trees INFORMS Journal on Optimization 1 164-183
[4]  
Bertsimas D(2017)Personalized diabetes management using electronic medical records Diabetes Care 40 210-217
[5]  
Dunn J(2001)Random forests Machine Learning 45 5-32
[6]  
Bertsimas D(2021)Mathematical optimization in classification and regression trees Top 29 5-33
[7]  
Dunn J(2010)Optimal constraint-based decision tree induction from itemset lattices Data Mining and Knowledge Discovery 21 9-51
[8]  
Mundru N(2018)Some methods for heterogeneous treatment effect estimation in high dimensions Statistics in Medicine 37 1767-1787
[9]  
Bertsimas D(2020)Learning optimal decision trees using constraint programming Constraints 25 226-250
[10]  
Kallus N(2018)Estimation and inference of heterogeneous treatment effects using random forests Journal of the American Statistical Association 113 1228-1242
1234