Co-creating a globally interpretable model with human inputs

Abstract

We consider an aggregated human-AI collaboration aimed at generating a joint interpretable model. The model takes the form of Boolean decision rules, where human inputs are provided in the form of logical conditions or as partial templates. This focus on the combined construction of a model offers a different perspective on joint decision making. Previous efforts typically have focused on aggregating decisions instead. We demonstrate the proposed approach through two examples and highlight usefulness and challenges of the approach.