Tim Jahn

Tim Jahn is a MATH+ junior research group leader for "Mathematics of Data Science" at the Technical University of Berlin since September 2023. Previously, he served as Acting Professor for "Data Assimilation" at the University of Potsdam from October 2023 to March 2024, and as a Research Associate with the "Hausdorff Postdoc" position at the University of Bonn from September 2021 to August 2023.

Dr. Jahn received his Doctorate with highest honors (summa cum laude) from the University of Frankfurt in April 2021, following an M.Sc. in Mathematics (2016, grade 1.0) and B.Sc. in Physics (2015, grade 1.1), both from the same institution. His academic journey included an Erasmus exchange at Stockholm University in 2014.

Dr. Jahn's research focuses on statistical inverse problems, stochastic optimization, dimension reduction, and analysis of neural networks. His work bridges theoretical mathematics with practical applications in data science, particularly in developing regularization methods that function without precise noise characteristics. His publication record shows a clear evolution from fundamental mathematical theory toward increasingly applied work connecting with machine learning methodologies, with his 2024 paper on "Early Stopping of Untrained Convolutional Neural Networks" representing a significant contribution at this intersection.

His research program has been supported through competitive fellowships including the prestigious Hausdorff Postdoc fellowship at the University of Bonn and his current MATH+ junior research group leadership at TU Berlin, funded by the Berlin Mathematics Research Center.

As leader of the "Mathematics of Data Science" group within the Institute of Mathematics, Dr. Jahn directs research that develops rigorous mathematical foundations for modern data analysis techniques, with particular emphasis on uncertainty quantification and the interface between classical inverse problems and contemporary machine learning approaches.