Nikita Kotelevskii

Graduated from: MSc Data Science, Skoltech

Research interests: Uncertainty quantification, Bayesian methods, Generative Modelling

Thesis title: Bayesian methods for uncertainty quantification

Biography:

Nikita Kotelevskii (born in Omsk, Russia on 11/08/1996) is a Data Scientist and a 2nd year Ph.D. student at the Skolkovo Institute of Science and Technology. He is currently working on Bayesian methods and uncertainty estimation. His tentative Ph.D. title is “Bayesian methods for uncertainty quantification.” Nikita completed his Bachelor’s degree in 2018 from Lomonosov Moscow State University, Physics department, Chair of Polymer and Crystal Physics. He was graduated with distinction. He received his MSc degree in Data Science in Skoltech 2020 under the supervision of Maxim Panov. His Master’s title was “Enhancing variational inference with Markov chain Monte Carlo”. During his MSc and Ph.D. studies, he published several papers, related to a combination of variational inference and Markov chain Monte Carlo methods. One of these papers, where specific applications to Variational Autoencoders were described, was published at ICML 2021.

Publications:

• A. Thin, N. Kotelevskii, A. Doucet, A. Durmus, E. Moulines, and M. Panov. Monte carlo variational auto-encoders. In International Conference on Machine Learning, pages 10247–10257. PMLR, 2021.
• N. Y. Kotelevskii, A. Fishkov, K. Fedyanin, A. Petiushko, and M. Panov. Nonparametric approach to uncertainty quantification for deterministic neural networks. In NeurIPS 2021 Workshop on Distribution Shifts: Connecting Methods and Applications, 2021.
• N. Y. Kotelevskii, A. Fishkov, K. Fedyanin, A. Petiushko, and M. Panov. Nuq: Nonparametric uncertainty quantification for deterministic neural networks. In ICML 2021 Workshop on Distribution-Free Uncertainty Quantification, 2021.
• A. Thin, N. Kotelevskii, J.-S. Denain, L. Grinsztajn, A. Durmus, M. Panov, and E. Moulines. Metflow: A new efficient method for bridging the gap between markov chain monte carlo and variational inference. arXiv preprint arXiv:2002.12253, 2020.

Academic Mobility:

I had an academic mobility experience at École Polytechnique, CMAP, under the supervision of Eric Moulines. First time September-December 2019. Second time October 2021 – January 2022.