Johannes Lederer

Johannes Lederer is Professor of Mathematics of Data-Based Methods at the University of Hamburg's Department of Mathematics. His research bridges mathematics, computer science, and applications, focusing on high-dimensional statistics, deep learning foundations, and robust machine learning.

Research interests cover theoretical aspects of deep learning including regularization techniques, statistical guarantees for neural networks, and adaptive methods for high-dimensional problems. Recent work establishes sample complexity requirements for deep networks and fairness benchmarks for computer vision algorithms.

Publications demonstrate interdisciplinary approaches, with developments in time-series forecasting, privacy-preserving methods, and geometric deep learning. Active collaborations include international workshops on statistics' role in modern AI.

Professional activities include editorial responsibilities and conference organization. Team management includes doctoral candidates and research groups exploring statistical learning and AI safety.