Julius Berner, B.Sc. M.Sc.
Educated at University of Vienna (BSc, MSc) with a specialization in applied mathematics and scientific computing, Julius Berner became very interested in machine learning and neural networks, in particular. His research focuses on the mathematical analysis of deep learning at the intersection of approximation theory, statistical learning theory, and optimization.
Pavol Harár, Ph.D.
Pavol Harár is a postdoctoral machine learning researcher and engineer focused on fundamental research in deep learning, specializing in learning from limited multi-modal medical data sets. He is skilled in machine hearing, computer vision, and visualization.
He received an MSc in System Engineering and Informatics and a PhD in Machine Learning from Brno University of Technology. He gained experience in predictive modeling, signal processing, and parallel computing as a member of Brain Diseases Analysis Laboratory and Numerical Harmonic Analysis Group. He joined the group in December 2019.
Lukas Liehr, B.Sc. M.Sc.
Lukas Liehr studied Mathematics at the Technical University of Munich and Seoul National University. He received his Master's degree with high distinction from the Technical University of Munich where Analysis and PDE were his selected focus of studies. Afterwards, he worked as a research assistant at Helmholtz Centre Munich. His current research focuses on the infinite-dimensional Gabor phase retrieval problem. He investigates this problem by using techniques from Functional Analysis and Harmonic Analysis. Moreover, he is interested in the design of stable algorithms for phase retrieval.
Dipl.-Ing. Michael Scherbela, M.Sc.
Michael Scherbela studied physics at the Graz University of Technology, specializing in computational physics. He joined the group in 2020 and is currently pursuing his PhD in mathematics, primarily conducting research on applications of deep learning methods to problems in the natural sciences. In particular he is interested in developing neural-network-based methods to find solutions to the Schrödinger equation, which lies at the heart of computational chemistry and quantum physics. To tackle this problem he combines existing quantum chemistry approaches with deep-neural networks, designs new network architectures for unsupervised learning, and optimizes existing approaches.