alexeizaitsev



Personal Websites

Lab site:

https://sites.skoltech.ru/larss/

Google Scholar profile:

https://scholar.google.ru/citations?user=WfzzY4cAAAAJ&hl=en

Alexey Zaytsev

Alexey has expertise in numerical methods, statistics, and industrial applications of data analysis.

In 2012, Alexey graduated from the Department of Control and Applied Mathematics of Moscow Institute of Physics and Technology (MIPT). In his Master’s thesis, Alexey proposed a modification of Bayesian approach for linear regression that allows an automated feature selection.

He then completed a Ph.D. in Math at IITP RAS in 2017. Alexey obtained a new result on the effectiveness of Bayesian procedures for Gaussian process regression and a first-ever theoretical justification for the selection of the design of experiments for variable fidelity models as well as minimax errors for Gaussian process regression. He published these results in a number of peer-reviewed journals and top-ranked conferences such as AISTATS.

During his studies at MIPT Alexey joined a Skolkovo resident company DATADVANCE and took part in the development of MACROS library dedicated to data analysis for engineers. He developed a pioneering industry-level tool for data fusion that solves a regression problem for the case of data with more than one fidelity. Alexey also completed several projects connected with the application of data analysis for such companies as AREVA, TOTAL, and Airbus.

In 2018 he was awarded Moscow government award for young scientists “Development of predictive analytics methods for processing industrial, biomedical and economic data” joint with E. Burnaev and M. Panov.

Now Alexey focuses his research on the development of new methods for Bayesian optimization and embeddings for weakly structured data. He also actively participates in industrial projects and teaching routines including joint projects with Sberbank and Gazprom Neft.

Development of

  • advanced algorithms for  Bayesian optimization
  • embeddings for weakly structured data
  • adversarial attacks for categorical data
  • anomaly detection approaches

Alexey also

  • leads Industrial projects for Sberbank, Huawei, and other major industrial companies,
  • develops and provides further education courses for industrial partners of Skoltech

Preprints:

  1. R. Kail, A. Zaytsev, E. Burnaev. Recurrent Convolutional Neural Networks help to predict locations of Earthquakes.
    arXiv preprint arXiv:2004.09140. 2020.
  2. I. Fursov, A. Zaytsev, N. Kluchnikov, A. Kravchenko, E. Burnaev. Gradient-based adversarial attacks on categorical sequence models via traversing an embedded world. arXiv preprint arXiv:2003.04173. 2020.
  3. S.C. Kumbhakar, A. Peresetsky, Y. Shchetynin, A. Zaytsev. Technical efficiency and inefficiency: Reassurance of standard SFA models and a misspecification problem. 2020.
  4. I. Fursov, A. Zaytsev, N. Kluchnikov, A. Kravchenko, E. Burnaev. Differentiable Language Model Adversarial Attacks on Categorical Sequence Classifiers. arXiv preprint arXiv:2006.11078. 2020
  5. I. Fursov, A. Zaytsev, R. Khasyanov, M. Spindler, E. Burnaev. Sequence embeddings help to identify fraudulent cases in healthcare insurance. arXiv preprint arXiv:1910.03072. 2020.

Published:

  1. V. Snorovikhina, A. Zaytsev. Unsupervised anomaly detection for discrete sequence healthcare data. arXiv preprint arXiv:2007.10098. AIST. 2020. 
  2. I. Fursov, A. Zaytsev, N. Kluchnikov, A. Kravchenko, E. Burnaev. Gradient-based adversarial attacks on categorical sequence models via traversing an embedded world. arXiv preprint arXiv:2003.04173. AIST. 2020
  3. E. Gurina, N. Klyuchnikov, A. Zaytsev et al. Application of machine learning to accidents detection at directional drilling. Journal of Petroleum Science and Engineering. 2020.
  4. E. Romanenkova, A. Zaytsev et al. Real-Time Data-Driven Detection of the Rock-Type Alteration During a Directional Drilling. IEEE Geoscience and Remote Sensing Letters. 2019.
  5. N. Klyuchnikov, A. Zaytsev et al. Data-driven model for the identification of the rock type at a drilling bit. Journal of Petroleum science and Engineering. 2019
  6. P. Proskura, A. Zaytsev, I. Braslavsky, E. Egorov, E. Burnaev.
    Usage of Multiple RTL Features for Earthquakes Prediction. ICMLA. 2019
  7. N. Kozlovskaya, A. Zaytsev. Deep Ensembles for Imbalanced Classification. ICMLA. 2017
  8. A. Zaytsev, E. Burnaev. Large scale variable fidelity surrogate modeling. Annals of Mathematics and Artificial Intelligence, 81: 167-186. 2017. https://link.springer.com/article/10.1007/s10472-017-9545-y
  9. A. Zaytsev, E. Burnaev. Minimax Approach to Variable Fidelity Data Interpolation. Artificial Intelligence and Statistics, 652-661. 2017. http://proceedings.mlr.press/v54/zaytsev17a.html
  10. A. Zaytsev. Variable Fidelity Regression Using Low Fidelity Function Blackbox and Sparsification. Symposium on Conformal and Probabilistic Prediction with Applications, 147-164. 2016. https://link.springer.com/chapter/10.1007/978-3-319-33395-3_11
  11. A. Zaytsev. Reliable surrogate modeling of engineering data with more than two levels of fidelity. Mechanical and Aerospace Engineering (ICMAE), 2016 7th International Conference. 2016. ieeexplore.ieee.org/abstract/document/7549563/
  12. A.A Zaitsev, E.V. Burnaev, V.G. Spokoiny. Properties of the posterior distribution of a regression model based on Gaussian random fields. Automation and Remote Control 74 (10), 1645-1655, 2013. link.springer.com/article/10.1134/S0005117913100056
  13. E.V. Burnaev, A.A. Zaytsev, V.G. Spokoiny (alphabetic order of authors). The Bernstein-von Mises theorem for regression based on Gaussian Processes. Russ. Math. Surv 68 (5), 954-956, 2013. iopscience.iop.org/article/10.1070/RM2013v068n05ABEH004863
  14. E.V. Burnaev, M.E. Panov, A.A. Zaytsev. Regression on the basis of nonstationary Gaussian processes with Bayesian regularization. Journal of Communications Technology & Electronics 61 (6), 661. 2016. https://link.springer.com/article/10.1134/S1064226916060061
  15. E.V. Burnaev, A.A. Zaytsev. Surrogate modeling of multifidelity data for large samples. Journal of Communications Technology & Electronics 60 (12), 1348. 2015. link.springer.com/article/10.1134/S1064226915120037
  16. A.A. Zaytsev, E.V. Burnaev, V.G. Spokoiny. Properties of the Bayesian parameter estimation of a regression based on Gaussian processes. J. Math. Sci 203 (6), 789-798. 2014. https://www.datadvance.net/assets/files/publications/zaytsev13fam.pdf

 

  • Masters in Math at MIPT
  • PhD in Statistics and Machine Learning at IITP RAS

Industrial companies:

  • Airbus
  • AREVA
  • TOTAL

Banks and related companies:

  • BPCE (top 3 bank in France)
  • Lending club
  • Sberbank

In 2018 Alexey was awarded the Moscow government award for young scientists “Development of predictive analytics methods for processing industrial, biomedical and economic data” joint with E. Burnaev and M. Panov.

Present students (with year of the planned defense)

  • 2025 Evgneiya Romanenkova, Skoltech, PhD
  • 2022 Matvey Morozov, Skoltech, Master
  • 2021 Nikita Bekezin, Skoltech, Master
  • 2021 Maria Begicheva, Skoltech, Master
  • 2021 Anna Nikolaeva, Skoltech, Master
  • 2021 Artem Zabolotny, Skoltech, Master
  • 2021 Polina Proscura, MIPT, Master
  • 2021 Roman Kail, MIPT, Bachelor

Past students

  • 2020 Ivan Fursov, Skoltech, Master, A (joint with E.Burnaev)
  • 2020 Viktoria Snorovikhina, Skoltech, Master, B
  • 2019 Evgneiya Romanenkova, IITP, Master, A
  • 2019 Vladimir Shenderov, Skoltech, Master, B (joint with L.Ismailova)
  • 2019 Rasul Khasianov, Skoltech, IITP, Master, B
  • 2018 Oleg Alenkin, Skoltech, Master, B. Bayesian Optimization of SHiP Elements (joint with A.
    Ustyuzhanin)
  • 2017 Natalya Kozlovskaya, HSE, Master, A. Deep Boosting for Imbalanced Classification (joint
    with E.Burnaev)
  • 2017 Lee Yunjeong, Skoltech, Master, B. Gas Consumption Forecast in China (joint with K.Letova)
  • 2020 Matvey Morozov, IITP, Bachelor, A
  • 2020 Timur Aminov, MIPT, Bachelor, A
  • 2019 Polina Proscura, IITP, Bachelor, A
  • 2014 Yermek Kapushev, MIPT, Bachelor, A. Multifidelity Gaussian Processes (joint with E.Burnaev)
  • Courses on sequential data processing (in progress) https://github.com/sk-larss/2020-sequencedata-lectures
  • Craft of data visualization https://github.com/likzet/craft_of_data_visualization