- Research Staff
Filip Svoboda is a researcher at the University of Cambridge, a visiting lecturer at the University of Oxford (AIMS), and a Teaching By-Fellow at the Churchill College, Cambridge. Prior to joining Cambridge he was part of the AIMS CDT program of the University of Oxford, and the Computational Statistics and Machine Learning MSc of the University College London.
Filip Svoboda’s research interests are in applied and federated deep learning focusing on the issues of efficiency and privacy. His work in the Space domain focused on superior data processing for remote sensing and on improved efficiency and deployment of distributed on-board learning. His was the first model to deliver competitive solar wind speed prediction based on EUV images of the sun. Also, his were some of the first applications of federated learning in orbit, both in simulation and on hardware. His work in medical, and pharmaceutical domains unlocked previously inaccessible data pools to the general research community and allowed researchers to combine insights from previously disjointed datasets. This was done by addressing the privacy, security, and data quality issues inherent in the domain and its regulatory environment. Finally, his team won the 2023 UK-US Prize for Privacy Tech Against Financial Crime.
He supervises part II and III theses in these domains as well as in a range of other applied data applications.
