David Penz

David Penz is a PreDoc Researcher at Technische Universität Wien (TU Wien), affiliated with the Faculty of Informatics and the Department of Machine Learning. He teaches courses such as Introduction to Machine Learning, Machine Learning Algorithms and Applications, and Theoretical Foundations and Research Topics in Machine Learning. His research focuses on advancing Graph Neural Networks (GNNs), particularly for outerplanar graphs, and developing ethical AI systems in recommendation engines. His work emphasizes data privacy, fairness in AI, and music information retrieval.

Key projects include the StruDL initiative (2023–2027), exploring maximally expressive GNN architectures. He has supervised a diploma thesis on antibody-antigen binding affinity prediction using geometric deep learning. Recent publications highlight contributions to unlearning sensitive user attributes in recommendations and emotion-aware music recommendation systems.

His academic output spans conferences like NeurIPS, SIGIR, and ICMR, demonstrating expertise in both theoretical and applied machine learning domains. He actively participates in academic challenges, such as the ACM Recommender Systems Challenge 2021, showcasing lightweight XGBoost approaches.