Multicriteria interpretability driven Deep Learning
This addresses the need for interpretable deep learning models in regulated fields like credit risk, though it appears incremental by building on existing interpretability constraints.
The authors tackled the problem of deep learning's lack of interpretability in high-stakes contexts by proposing a multicriteria technique that injects knowledge into the objective function to control feature effects from the start, resulting in performant and robust models that overcome biases from data scarcity.
Deep Learning methods are renowned for their performances, yet their lack of interpretability prevents them from high-stakes contexts. Recent model agnostic methods address this problem by providing post-hoc interpretability methods by reverse-engineering the model's inner workings. However, in many regulated fields, interpretability should be kept in mind from the start, which means that post-hoc methods are valid only as a sanity check after model training. Interpretability from the start, in an abstract setting, means posing a set of soft constraints on the model's behavior by injecting knowledge and annihilating possible biases. We propose a Multicriteria technique that allows to control the feature effects on the model's outcome by injecting knowledge in the objective function. We then extend the technique by including a non-linear knowledge function to account for more complex effects and local lack of knowledge. The result is a Deep Learning model that embodies interpretability from the start and aligns with the recent regulations. A practical empirical example based on credit risk, suggests that our approach creates performant yet robust models capable of overcoming biases derived from data scarcity.