Rebekka Burkholz

Dr. Rebekka Burkholz is a tenured faculty member at the CISPA Helmholtz Center for Information Security in Saarbrücken, Germany, leading the Relational Machine Learning Group. Her research bridges machine learning and complex network science to develop robust, data-efficient models with applications in molecular biology. Previously, she held positions at Harvard T.H. Chan School of Public Health and ETH Zurich.

• PhD in Systems Design (2016) from ETH Risk Center
• Mathematics and Physics BSc/MSc from TU Darmstadt

Her work focuses on sparse training methods and theoretical deep learning, addressing challenges like computational efficiency and adversarial robustness. Recent publications explore:

• Sparse training via implicit sparsification
• GNN optimization through rescaling and rewiring
• Integration of domain knowledge in biomedical modeling
• Theoretical guarantees for batch normalization and lottery tickets

Scientific awards include:

• Zurich Dissertation Prize (2016)
• CSF Best Contribution Award (2016)

She actively advises PhD students and collaborates with interdisciplinary teams in biostatistics and systems biology.