We have extended the well-known anonymization tool ARX for biomedical data with machine learning techniques to support the creation of privacy-preserving machine learning models. With the tool presented in this article, accurate models can be created that preserve the privacy of patients and probands represented in the training set in a variety of threat scenarios. Our implementation is available as open source software. This work was carried out in collaboration with the Institute of Medical Informatics, Statistics and Epidemiology at Technical University of Munich. The complete article is available here.