Central University

Stanford-Elsevier Top 2% Scientists. A leading Russian expert in intelligent sensing, IoT, and behavioral/motor pattern analysis.

Since 2011, Vice Dean for Research at the Faculty of Bioengineering and Bioinformatics; since 2016, Deputy Director for Research at the Belozersky Institute of Physico-Chemical Biology, MSU. Specialist in microbial bioenergetics and ATP synthase function.

Statistics, decision theory, high-dimensional data analysis, and mathematical methods of machine learning

01 High-Dimensional Statistics

This direction focuses on developing and applying modern statistical methods for data characterized by a large number of features. Key tasks include estimating covariance and precision matrices under high dimensionality, which is critical for dependency analysis and building graphical models. We investigate clustering methods adapted for complex high-dimensional structures, as well as logistic regression using both frequentist and Bayesian approaches for classification problems. Special attention is given to nonparametric methods, such as probability density estimation and regression with instrumental variables, enabling the modeling of complex distributions that cannot be described by standard parametric families.

02 Statistics for Complex Structured Models

This direction focuses on data with complex internal organization. It includes the development of single-index and multi-index models, as well as manifold learning methods that effectively reduce dimensionality while preserving the nonlinear structure of information. We actively work with panel data, applying models with individual and time effects to analyze process dynamics. An important part of our research involves signal recovery methods and multivariate density estimation in problems with specific structural constraints. This direction also covers the integration of machine learning methods into classical and modern financial mathematics.

03 Optimization and Machine Learning

This direction combines theoretical developments in optimization and machine learning algorithms with their practical applications in finance. We investigate efficient training methods for multi-class classification models and algorithms for solving minimax problems using stochastic optimization. Significant attention is given to reinforcement learning for optimal portfolio management, hedging of derivative financial instruments with transaction costs, and optimal trade execution. For financial market modeling and risk management, we develop and apply generative models (including variational autoencoders), neural operators, and complex stochastic volatility models.

Researcher in the fields of physics, chemistry, bioinformatics, and machine learning. Author of over 70 scientific publications, 39 of which are in Q1/Q2 journals, including 15 Q1 articles over the past three years

A leading Russian expert in control theory, automation, and reinforcement learning with formal guarantees.

A leading Russian scientist working at the intersection of biotechnology, medical biochemistry, physical chemistry, and pharmacology