Bio
Vitaly conducted research as a PhD student under the supervision of Professor Kody Law from the University of Manchester. Prior joining the Turing doctoral programme he obtained BSc and MSc in Applied Mathematics and Physics from the Moscow Institute of Physics and Technology (MIPT), also he benefited by finishing the Data Science MSc track from the Skolkovo Institute of Science and Technology (Skoltech).
Vitaly has industrial research experience resulted in a number of R&D projects in mathematical modelling of wireless telecommunication networks, aspiring to combine and adapt the cutting-edge research breakthroughs with the industry needs.
Research interests
The main outcomes of Vitaly's research project revolve around developing scalable, efficient, and flexible modelling techniques in Bayesian inference and machine learning. In one study, they proposed an economical and scalable alternative to exact Bayesian inference in a regression problem with sparsity-promoting priors, which successfully handles larger datasets while maintaining comparable performance in variable selection and uncertainty quantification. In another study, they developed a method combining deep neural networks and mixtures of sparse Gaussian processes experts, providing a flexible, robust model that effectively handles complex data behaviours while maintaining superior performance in terms of accuracy and uncertainty quantification. Finally, they combined Bayesian inference with sparse system identification for model evaluation and selection in the context of tokamak plasma boundary dynamics simulation.
Selected publications and papers
'Clement Etienam, Kody Law, Sara Wade, and Vitaly Zankin. "Fast Deep Mixtures of Gaussian Process Experts". arXiv:2006.13309 (2022)
Sebastian De Pascuale, Vitaly Zankin, Jeremy D. Lore, Ben Russo, Paul Laiu, Birdy Phathanapirom, Steven L. Brunton, J. Nathan Kutz. "Uncertainty Quantification for Model Predictive Control of Tokamak Plasma Boundary Simulations with SOLPS-ITER". 64th Annual Meeting of the APS Division of Plasma Physics (2022)
Vitaly Zankin, Kody Law. "Sparse Online Variational Bayesian Inference". SIAM Conference on Uncertainty Quantification (2022)
Kody Law, Vitaly Zankin. "Sparse Online Variational Bayesian Regression". SIAM/ASA Journal on Uncertainty Quantification (2022)
Vitaly Zankin, Kody Law. "Sparse Online Inference with UQ for Regression and Inverse Problems". SIAM Conference on Computational Science and Engineering (2021)