Dr. rer. nat. Vladyslav Fediukov
Technical University of Munich
TUM School of CIT
Department of Computer Science
Boltzmannstrasse 3
85748 Garching
Germany
Office: MI 02.05.040
Mail: vladyslav.fediukov (at) tum.de
Office Hours: by arrangement
Links: Google Scholar
Background
- Started my postdoc @ SCML in November 2025
- Ph.D. topic “Multi-fidelity Gaussian processes for terramechanical modeling”, a joint project between TUM and DLR within the MUDS programm, October 2020 - September 2025
- M.Sc. in Computer science @ University of Tartu, September 2017 - March 2020
- B.Sc. in System analysis @ Igor Sikorsky Kyiv Polytechnic Institute, September 2013 - June 2017
Research interests
In general, Scientific Machine Learning. In particular, I am interested in:
- Fast/Alternative training methods for neural networks
- Understanding Machine Learning Methods and how to incorporate prior knowledge
- Bridging the gap between Neural networks and kernel methods
- Applications to physical simulations or engineering problems
- Incorporating uncertainties in the training and prediction process
Right now, I am working on:
- input/measurements uncertainty propagation in multi-fidelity Gaussian processes
Thesis Supervision
I can supervise students from across CIT. If you are interested in doing a thesis or research project with me (or other members of the group), please apply through the Thesis Application Form of the group or the thesis portal for the listed topics. You are invited to also propose your own topic in addition to the listed ones.
In case of any questions, feel free to also send me an e-mail.