Demonstrating non-inferiority of easy interpretable methods for insolvency prediction

In our case study, interpretable hypotheses perform better than other hypotheses. DTs have a better β-error, whereas DNFs outperform DTs with respect to the α-error. We compared both hypothesis classes and exemplary hypotheses. Both are interpretable in different ways leaving the choice to preference. This leads to the conclusion that interpretable threshold based methods are appropriate for classification problems in finance. In this domain, they are not inferior to more sophisticated methods like SVMs.