About the Laboratory

The laboratory unites domestic and foreign researchers working at the intersection of modern mathematical statistics, optimization, probability theory, theory of algorithms and other mathematical disciplines. The main goal of the laboratory is to develop new probabilistic and statistical approaches to solve urgent problems arising in modern data analysis.

The laboratory's activities are aimed at creating a unified approach to a wide range of tasks in which the complexity of data requires the use of new effective statistical approaches and algorithms. In particular, the central research topics are:

 1. Reinforcement learning 
 2. Generative modeling, including using modern deep learning models (generative adversarial networks, variational autoencoders, diffusion models) 
 3. Uncertainty analysis, reduction of variance, acceleration of MCMC algorithms
 4. Uncertainty analysis in the estimation of high-dimensional covariance matrices and their spectral projectors, applications to effective dimensionality reduction 
 5. Data analysis using the geometry of the Monge-Kantorovich space
 6. Estimation of discretely observed stochastic processes
 7. Statistical estimation of the McKean-Vlasov equations 
 8. An important component of the laboratory's activities is the education of young Russian world-class scientists.

Cooperation:

Our laboratory cooperates with researchers from universities and leading research centers such as: