Winfried Ripken

Winfried Ripken is a PhD candidate and research associate at Technical University of Berlin, affiliated with BIFOLD (Berlin Institute for the Foundations of Learning and Data). He pursues doctoral studies in Computer Science under Stefan Chmiela's supervision and participates in the BIFOLD graduate school.

His academic background includes:

• M.Sc. in Computer Science from Hasso Plattner Institute (2021)
• B.Sc. in Computer Science from TU Chemnitz (2017)

Ripken specializes in applying machine learning to natural sciences, with core expertise in Graph Neural Networks, Geometric Deep Learning, and Quantum Chemistry. His research bridges scientific computing and deep learning to solve complex problems like boundary value equations and molecular simulations, emphasizing interpretability and operator learning.

Recent publications reveal a strong trend toward scientific machine learning applications, particularly in physics-informed neural networks and image processing. His work demonstrates consistent innovation in adapting graph-based architectures for scientific domains while advancing whitebox methodologies in computer vision.

As a doctoral researcher, Ripken has no supervisees but contributes significantly through active GitHub development (26 repositories, 141 stars) and prior industry experience. He previously worked two years as a full-time machine learning researcher at Merantix Momentum, focusing on Graph Neural Networks and interpretability.

He operates within Stefan Chmiela's research group at BIFOLD, collaborating on interdisciplinary projects that integrate machine learning with chemistry, physics, and engineering challenges through Berlin's prominent AI research ecosystem.