Transparency and Predicive Power: Explaining Complex Classification Models

Complex classification models like neural networks usually have lower errors than simple models. They often have very many interdependent parameters, whose effects no longer can be understood by the user. For many applications, especially in the financial industry, it is vital to understand the reasons why a classification model arrives at a specific decision. We propose to use the full model for the classification and explain its predictive distribution by an explanation model capturing its main functionality. For a real world credit scoring application we investigate a spectrum of explanation models of different type and complexity.