POETREE: Interpretable Policy Learning with Adaptive Decision Trees
This work addresses the need for interpretable policy learning in clinical decision support systems, offering a novel method that is incremental in improving upon existing approaches focused solely on performance.
The paper tackles the problem of learning interpretable models of human decision-making from observed behavior, specifically in clinical care, by introducing POETREE, a framework that builds probabilistic tree policies and outperforms state-of-the-art methods on real and synthetic medical datasets in both interpretability and imitation accuracy.
Building models of human decision-making from observed behaviour is critical to better understand, diagnose and support real-world policies such as clinical care. As established policy learning approaches remain focused on imitation performance, they fall short of explaining the demonstrated decision-making process. Policy Extraction through decision Trees (POETREE) is a novel framework for interpretable policy learning, compatible with fully-offline and partially-observable clinical decision environments -- and builds probabilistic tree policies determining physician actions based on patients' observations and medical history. Fully-differentiable tree architectures are grown incrementally during optimization to adapt their complexity to the modelling task, and learn a representation of patient history through recurrence, resulting in decision tree policies that adapt over time with patient information. This policy learning method outperforms the state-of-the-art on real and synthetic medical datasets, both in terms of understanding, quantifying and evaluating observed behaviour as well as in accurately replicating it -- with potential to improve future decision support systems.