Department of Computer Science


The Data Science & Analytics Research (DSAR) group focuses on the qualitative assessment, interpretability, robustness and fairness of machine learning methods and their scalable applicability in real-life scenarios.


A Gradient-Based Split Criterion for Highly Accurate and Transparent Model Trees by Dr. Klaus Broelemann and Dr. Gjergji Kasneci was accepted at IJCAI 2019

  • The work presents a new split-criterion to boost interpretability of powerful tree based algorithms