Contextualized Policy Recovery: Modeling and Interpreting Medical Decisions with Adaptive Imitation Learning

Jannik Deuschel, Caleb Ellington, Yingtao Luo, Ben Lengerich, Pascal Friederich, Eric P. Xing
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:10642-10660, 2024.

Abstract

Interpretable policy learning seeks to estimate intelligible decision policies from observed actions; however, existing models force a tradeoff between accuracy and interpretability, limiting data-driven interpretations of human decision-making processes. Fundamentally, existing approaches are burdened by this tradeoff because they represent the underlying decision process as a universal policy, when in fact human decisions are dynamic and can change drastically under different contexts. Thus, we develop Contextualized Policy Recovery (CPR), which re-frames the problem of modeling complex decision processes as a multi-task learning problem, where each context poses a unique task and complex decision policies can be constructed piece-wise from many simple context-specific policies. CPR models each context-specific policy as a linear map, and generates new policy models on-demand as contexts are updated with new observations. We provide two flavors of the CPR framework: one focusing on exact local interpretability, and one retaining full global interpretability. We assess CPR through studies on simulated and real data, achieving state-of-the-art performance on predicting antibiotic prescription in intensive care units ($+22$% AUROC vs. previous SOTA) and predicting MRI prescription for Alzheimer’s patients ($+7.7$% AUROC vs. previous SOTA). With this improvement, CPR closes the accuracy gap between interpretable and black-box methods, allowing high-resolution exploration and analysis of context-specific decision models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-deuschel24a, title = {Contextualized Policy Recovery: Modeling and Interpreting Medical Decisions with Adaptive Imitation Learning}, author = {Deuschel, Jannik and Ellington, Caleb and Luo, Yingtao and Lengerich, Ben and Friederich, Pascal and Xing, Eric P.}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {10642--10660}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/deuschel24a/deuschel24a.pdf}, url = {https://proceedings.mlr.press/v235/deuschel24a.html}, abstract = {Interpretable policy learning seeks to estimate intelligible decision policies from observed actions; however, existing models force a tradeoff between accuracy and interpretability, limiting data-driven interpretations of human decision-making processes. Fundamentally, existing approaches are burdened by this tradeoff because they represent the underlying decision process as a universal policy, when in fact human decisions are dynamic and can change drastically under different contexts. Thus, we develop Contextualized Policy Recovery (CPR), which re-frames the problem of modeling complex decision processes as a multi-task learning problem, where each context poses a unique task and complex decision policies can be constructed piece-wise from many simple context-specific policies. CPR models each context-specific policy as a linear map, and generates new policy models on-demand as contexts are updated with new observations. We provide two flavors of the CPR framework: one focusing on exact local interpretability, and one retaining full global interpretability. We assess CPR through studies on simulated and real data, achieving state-of-the-art performance on predicting antibiotic prescription in intensive care units ($+22$% AUROC vs. previous SOTA) and predicting MRI prescription for Alzheimer’s patients ($+7.7$% AUROC vs. previous SOTA). With this improvement, CPR closes the accuracy gap between interpretable and black-box methods, allowing high-resolution exploration and analysis of context-specific decision models.} }
Endnote
%0 Conference Paper %T Contextualized Policy Recovery: Modeling and Interpreting Medical Decisions with Adaptive Imitation Learning %A Jannik Deuschel %A Caleb Ellington %A Yingtao Luo %A Ben Lengerich %A Pascal Friederich %A Eric P. Xing %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-deuschel24a %I PMLR %P 10642--10660 %U https://proceedings.mlr.press/v235/deuschel24a.html %V 235 %X Interpretable policy learning seeks to estimate intelligible decision policies from observed actions; however, existing models force a tradeoff between accuracy and interpretability, limiting data-driven interpretations of human decision-making processes. Fundamentally, existing approaches are burdened by this tradeoff because they represent the underlying decision process as a universal policy, when in fact human decisions are dynamic and can change drastically under different contexts. Thus, we develop Contextualized Policy Recovery (CPR), which re-frames the problem of modeling complex decision processes as a multi-task learning problem, where each context poses a unique task and complex decision policies can be constructed piece-wise from many simple context-specific policies. CPR models each context-specific policy as a linear map, and generates new policy models on-demand as contexts are updated with new observations. We provide two flavors of the CPR framework: one focusing on exact local interpretability, and one retaining full global interpretability. We assess CPR through studies on simulated and real data, achieving state-of-the-art performance on predicting antibiotic prescription in intensive care units ($+22$% AUROC vs. previous SOTA) and predicting MRI prescription for Alzheimer’s patients ($+7.7$% AUROC vs. previous SOTA). With this improvement, CPR closes the accuracy gap between interpretable and black-box methods, allowing high-resolution exploration and analysis of context-specific decision models.
APA
Deuschel, J., Ellington, C., Luo, Y., Lengerich, B., Friederich, P. & Xing, E.P.. (2024). Contextualized Policy Recovery: Modeling and Interpreting Medical Decisions with Adaptive Imitation Learning. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:10642-10660 Available from https://proceedings.mlr.press/v235/deuschel24a.html.

Related Material