Beyond Sparsity: Tree Regularization of Deep Models for Interpretability
This work addresses the interpretability barrier for deep models in critical applications like healthcare, though it is incremental as it builds on existing regularization methods.
The authors tackled the problem of deep model interpretability by introducing tree regularization to make predictions more easily simulated by humans, achieving comparable predictive accuracy to L1/L2 penalties while improving interpretability on medical tasks like sepsis and HIV treatment.
The lack of interpretability remains a key barrier to the adoption of deep models in many applications. In this work, we explicitly regularize deep models so human users might step through the process behind their predictions in little time. Specifically, we train deep time-series models so their class-probability predictions have high accuracy while being closely modeled by decision trees with few nodes. Using intuitive toy examples as well as medical tasks for treating sepsis and HIV, we demonstrate that this new tree regularization yields models that are easier for humans to simulate than simpler L1 or L2 penalties without sacrificing predictive power.