Jakob is a PhD student in the Department of Mathematics at King’s College London, where he works on the reliability, computability, and geometry of AI. He currently serves as a teaching assistant for the course Statistical Modelling, Machine Learning and Artificial Intelligence.

He completed his BSc and MSc in Mathematics at the University of Würzburg. His master’s thesis, The Geometric Whitney Problem and Approximations of Neural Networks on Manifolds, supervised by Knut Hüper, studied connections between geometry and AI. His bachelor’s thesis, The Arens–Royden Theorem, supervised by Oliver Roth and Stefan Waldmann, involved complex analysis and functional analysis.

During his MSc studies, Jakob undertook two research visits to King’s College London and completed student exchanges at Ajou University in South Korea and Aristotle University of Thessaloniki in Greece. He also worked as a student researcher funded by the BMBF project DyCA.

Jakob organises the research seminar Mathematics of Machine Learning for the Elite Network of Bavaria.

Research interests

• Trustworthiness and reliability of AI
• Robustness and stability of AI
• Computational barriers in Optimization
• Linear Programming
• Geometry of data
• Manifold Hypothesis
• Approximation theory of neural networks
• High-dimensional learning

Thesis title

On the reliability of AI: robustness and computational barriers

PhD supervision

First supervisor: Dr Alexander Bastounis

Second supervisor: Dr Spyridon Pougkakiotis