Models of stochastic dynamics

  • 基本信息
其他名称: 24-71-10109
项目负责人: Zhitlukhin Mikhail
发表日期: 2024
主持机构: Steklov Mathematical Institute of Russian Academy of Sciences,
国家: 俄罗斯
开始日期: 2024
结束日期: 2027
简介: Models of stochastic dynamics are central to modern probability theory and arise in many areas, including stochastic process theory, statistical sequential analysis, stochastic analysis, dynamical systems theory, and random matrix theory. The analysis of such models plays a fundamental role for applications including financial mathematics, stochastic game theory, etc. The main models of stochastic dynamics and related research directions of the project are:$One of the important examples of the application of models with stochastic dynamics is the theory of statistical sequential analysis. Problems from this field consist in estimating some unobserved characteristics from an observed random process, and the prominent feature of estimation procedures is their dynamic construction (i.e. iterative in time). In the project we will consider new settings of statistical sequential analysis problems for a Bayesian model of Brownian motion with drift (i.e., assuming that the magnitude of drift is unknown but the distribution of its possible values is known).$Models of Brownian motion with drift are among the most popular basic models in applications; however, the case of nonlinear drift has not been sufficiently studied. The project separately intends to investigate the power drift occurring at a random time, in the case when the power index takes a critical value (at the boundary between singularity and absolute continuity of the distribution of the process with drift with respect to the Wiener measure). This case has a close connection with Gaussian multiplicative chaos.$The next model of stochastic dynamics we plan to investigate in the project will be branching random walks. Currently, methods based on spectral operator theory are actively developed in the world to analyze the asymptotic properties of branching random walks. The theory of orthogonal polynomials and Jacobi matrices allows us to study some models of branching random walks without the assumption of spatial homogeneity of the underlying walk. These methods have a deep connection with the methods used to analyse multiplicative chaos and based on spectral operator theory.$The study of random measures and multiplicative chaos is one of the most actively developed area in random processes theory. We are planning to study the multiplicative chaos for random entire functions corresponding to determinantal point processes arising in the random matrix theory.$The project also intends to investigate random matrices whose rows form a time series. Despite active research in random matrix theory, which is important for many applications in statistics, machine learning, physics, etc., questions about the limiting behavior of the singular spectrum (the set of singular values) of random matrices with nonlinear dependencies remain largely open.$The last research area of the project will be stochastic dynamics models related to evolutionary games. The plan is to explore different classes of evolutionarily optimal strategies and how they determine the structure of the game.
专业领域: 其他
语种: 英语

相关项目

Models of stochastic dynamics

Models of stoch... 2024