Lie algebra & group methods to non-linear stochastic dynamical systems and its applications

点击翻译按钮获取中文摘要

The present study focuses on Lie algebra & group methods to formulate numerical schemes of stochastic differential equations and the related topics on non-linear stochastic dynamical systems governed by such stochastic equations.The head investigator, Misawa, deeply investigates "symplectic integrators" for stochastic Hamiltonian dynamical systems by using "composition methods"which are formulated on the basis of Lie algebra & group theory. The new schemes are advantageous to preserve symplectic structure of the stochastic Hamiltonian systems numerically and are useful for producing the stable numerical solutions. To examine the superiority, Misawa also gives an illustrative example on the proposed schemes for such stochastic systems.In order to investigate the utility of stochastic numerical analysis, the investigators study on non-linear stochastic dynamical systems in various research fields. Misawa studies numerical simulations of stochastic macroeconomic models, Markov chain algor … More ithm in smoothing time series data and stochastic and statistical models in financial engineering. Particularly, he mainly works with investigating some regression models concerning with electricity markets. The investigator, Miyahara, studies the option pricing problems in the incomplete asset market, which is one of the important problems in the field of mathematical finance. He also constructs the [Geometric Levy process & MEMM] pricing model, in which the geometric Levy processes are adopted as the underlying asset price processes and the MEMM(=minimal entropy martingale measure), and investigates the properties of this model. The investigator, Shimizu, works with the following topics ; coalescent process with fluctuating population size and its effective population size, Poisson random measures and infinitely divisible distributions in population genetics, and the Handa model incorporating symmetric selections. The investigator, Notohara studies the genealogy of sampled genes from geographically structured population which is described by a Markov process called Structured Coalescent Model. He investigates effects of the geographical structure on the distribution of the coalescence time, which is the time to the most recent common ancestor of samples, and the distribution of the number of segregating sites detected in sampled DNA sequences. The investigator, Hashimoto, studies non-isotropic Gevrey hypoellipticity for Grushin operators and the analyticity of the solutions of the non-linear elliptic partial differential equations. Through these related topics, we find out that stochastic dynamical theory and stochastic numerical methods are useful for the analysis of the several stochastic models. Less