معرفی
Prof. Dr. Philipp Grohs is a full professor of Mathematical Data Science at the University of Vienna and head of the Mathematical Data Science group at RICAM (Austrian Academy of Sciences). He holds a MSc from TU Vienna (2006) and a PhD from the same institution (2007). After postdoc positions at TU Graz, KAUST, and ETH Zurich, he became an assistant professor at ETH Zurich in 2011 before moving to the University of Vienna in 2016. His research focuses on designing efficient algorithms for signal processing, computational sciences, and finance, with recent contributions to understanding deep learning algorithms and solving high-dimensional PDEs using machine learning. He has received the ETH Latsis Prize (2014) and was selected for an Alexander von Humboldt Professorship (2019).
Research Interests:
- Mathematical foundations of deep learning
- High-dimensional PDEs and their numerical solutions
- Signal and image processing (phase retrieval, Gabor systems)
- Approximation theory and function spaces
- Computational finance and mathematical modeling
Recent Research Trends: His work explores theoretical guarantees for neural network performance, particularly in overcoming dimensionality challenges. Notable contributions include phase retrieval algorithms, analysis of DNN expressivity, and applications of deep learning in quantum chemistry and epidemiology modeling.
Awards:
- ETH Latsis Prize (2014)
- Alexander von Humboldt Professorship (2019)
Advising & Projects: Leads research initiatives such as the 'Data Science' hub at Vienna, and has coordinated projects on hybrid computational sciences and explainable AI models. Active in supervising interdisciplinary research teams across mathematics and computer science.
Labs/Teams: Directs the Mathematical Data Science group at RICAM and oversees computational research collaborations with institutions like KAUST and ETH Zurich. Engages in applied projects like group testing strategies for SARS-CoV-2 and neural network-based electronic structure calculations.
