Philip Naumann
Philip Naumann
Doctoral Researcher
Affiliation: BIFOLD
Philip Naumann is a PhD student in Machine Learning at the Berlin Institute for the Foundations of Learning and Data (BIFOLD) at Technische Universität Berlin. His research explores how distribution shifts can be explained and modeled through the lens of optimal transport and XAI, with the goal of improving the reliability of foundation models. His work also extends to medical and industrial applications. He holds an MSc in Computer Science from Leibniz Universität Hannover, where he also spent an exchange semester at The Chinese University of Hong Kong, and has held research positions at Freie Universität Berlin and the L3S Research Center.
- Optimal Transport
- Explainable AI
- Evolutionary Algorithms
Philip Naumann, Jacob Kauffmann, Klaus-Robert Müller, Grégoire Montavon
Reliable Modeling of Distribution Shifts via Displacement-Reshaped Optimal Transport
Philipp Wissmann, Philip Naumann, Daniel Hein, Steffen Udluft, Marc Weber, Simon Leszek, Thomas Runkler
Efficient and Resilient Machine Learning for Industrial Applications
Jonah Kömen, Edwin D. de Jong, Julius Hense, Hannah Marienwald, Jonas Dippel, Philip Naumann, Eric Marcus, Lukas Ruff, Maximilian Alber, Jonas Teuwen, Frederick Klauschen, Klaus-Robert Müller
Towards Robust Foundation Models for Digital Pathology
Philip Naumann, Jacob Kauffmann, Grégoire Montavon
Wasserstein Distances Made Explainable: Insights into Dataset Shifts and Transport Phenomena
Philip Naumann
Towards XAI for Optimal Transport
Wasserstein distances made explainable
BIFOLD scientists developed a novel framework to make a widely used foundational statistical tool, the Wasserstein distance, interpretable in machine learning and data analysis contexts.