The mathematical-statistical methods that SCHUFA currently uses for credit scoring are logistic regression and gradient tree boosting. Both methods learn on the basis of examples under human guidance by independently extracting statistical correlations from the data.

We train the procedures with reliable data and rely on experts who check and adjust the systems with regard to prediction quality, reliability and stability. This means that people at SCHUFA are always in control. The scientific nature and selectivity of the procedures used have been confirmed by independent external experts and the procedures have been disclosed to the responsible data protection supervisory authority.

The procedures are based on the personal data that is also shown in the in accordance with Art. 15 GDPR.