About me
News: In July 2023, I will start my next postdoc position with Samuel Fiorini at Université Libre de Bruxelles.
I am a postdoctoral researcher in combinatorial optimization and theory of artificial neural networks at the LSE math department, working with László Végh and funded by his ERC grant ScaleOpt. Between 2018 and 2022, I completed my PhD at TU Berlin, supervised by Martin Skutella and funded by the graduate school Facets of Complexity. I obtained my BSc and MSc degrees in Mathematics from TU Kaiserslautern under the supervision of Sven O. Krumke with scholarships from the German Academic Scholarship Foundation (Studienstiftung des deutschen Volkes) and the Felix-Klein-Zentrum für Mathematik. For my master thesis, I received an award by the German Operations Research Society (GOR). In 2015, I spent one semester at the National University of Singapore. I qualified for the International Mathematical Olympiad (IMO) 2013, but did not participate because I spent the summer at the International Summer Science Institute on the campus of the Weizmann Institute in Rehovot, Israel.
My brother Johannes works in mathematical image processing.
Drop me an email if you want to talk about combinatorics of neural networks or join a bouldering session!
Accepted Publications
Towards Lower Bounds on the Depth of ReLU Neural Networks
with Amitabh Basu, Marco Di Summa, and Martin Skutella
Accepted for SIAM Journal on Discrete Mathematics (SIDMA 2023)
Conference version at Conference on Neural Information Processing Systems (NeurIPS 2021)
Conference Version |
Preprint |
Video
Provably Good Solutions to the Knapsack Problem via Neural Networks of Bounded Size
with Martin Skutella
INFORMS Journal on Computing (IJOC 2023)
Conference Version at AAAI 2021 Conference on Artificial Intelligence
Journal Version |
Conference Version |
Preprint |
Video
ReLU Neural Networks of Polynomial Size for Exact Maximum Flow Computation
with Leon Sering
Conference on Integer Programming and Combinatorial Optimization (IPCO 2023)
Conference Version |
Preprint
Lower Bounds on the Depth of Integral ReLU Neural Networks via Lattice Polytopes
with Christian Haase and Georg Loho
International Conference on Learning Representations (ICLR 2023)
Conference Version |
Preprint
The Computational Complexity of ReLU Network Training Parameterized by Data Dimensionality
with Vincent Froese and Rolf Niedermeier
Journal of Artificial Intelligence Research (JAIR 2022)
Journal Version |
Preprint
Online Algorithms to Schedule a Proportionate Flexible Flow Shop of Batching Machines
with Christian Weiß, Heiner Ackermann, Sandy Heydrich, and Sven O. Krumke
Journal of Scheduling (J Sched 2022)
Journal Version |
Preprint
Coloring Drawings of Graphs
with Felix Schröder and Raphael Steiner
Electronic Journal of Combinatorics (E-JC 2022)
Journal Version |
Preprint
Scheduling a Proportionate Flow Shop of Batching Machines
with Christian Weiß, Heiner Ackermann, Sandy Heydrich, and Sven O. Krumke
Journal of Scheduling (J Sched 2020)
Journal Version |
Preprint
Decision Support for Material Procurement
with Heiner Ackermann, Erik Diessel, Michael Helmling, Neil Jami, and Johanna Schneider
Operations Research Proceedings 2019
Conference Version
Sweep Algorithms for the Capacitated Vehicle Routing Problem with Structured Time Windows
with Philipp Hungerländer and Christian Truden
Operations Research Proceedings 2018
Conference Version |
Preprint
Preprints
Mode Connectivity in Auction Design
with Yixin Tao and László A. Végh (2023)
Preprint
Training Neural Networks is NP-Hard in Fixed Dimension
with Vincent Froese (2023)
Preprint
Training Fully Connected Neural Networks is ∃R-Complete
with Daniel Bertschinger, Paul Jungeblut, Tillmann Miltzow, and Simon Weber (2022)
Preprint
Theses
Facets of Neural Network Complexity
PhD Thesis (2022)
Scheduling a Proportionate Flow Shop of Batching Machines
Master Thesis (2018)
Received an award from the German Operations Research Society (GOR)
Summary published in OR Proceedings 2019
A Parametric View on Robust Graph Problems
Bachelor Thesis (2016)