188比分直播

Mean-square approximations of Levy noise driven SDEs with super-linearly growing diffusion and jump

发布者:文明办发布时间:2020-11-06浏览次数:213


主讲人:甘四清  中南大学教授


时间:2020年11月7日14:00


地点:腾讯会议 533 332 192


举办单位:数理学院


主讲人介绍:甘四清,博士,中南大学教授,博士生导师,2001年毕业于中国科学院数学研究所,获理学博士学位,2001-2003年在清华大学计算机科学与技术系高性能计算研究所做博士后,曾先后访问美国、新加坡、香港等科研院所。主要研究方向为确定性微分方程和随机微分方程数值解法。主持国家自然科学基金面上项目4项,  参加国家自然科学基金重大研究计划集成项目1项。在《SIAM Journal on Scientific Computing》、 《BIT Numerical  Analysis》、《Applied Numerical Mathematics》、《Journal of Mathematics Analysis and  Applications》、《中国科学》等国内外学术刊物上发表论文90余篇。2005年入选湖南省首批新世纪121人才工程。2014年湖南省优秀博士学位论文指导老师。  


内容介绍:We first establish a fundamental mean-square convergence theorem for general  one-step numerical approximations of Levy noise driven stochastic differential  equations with non-globally Lipschitz coefficients. Then two novel explicit  schemes are designed and their convergence rates are exactly identified via the  fundamental theorem. Different from existing works, we do not impose a globally  Lipschitz condition on the jump coefficient but formulate appropriate  assumptions to allow for its super-linear growth. However, we require that the  Levy measure is finite. New arguments are developed to handle essential  difficulties in the convergence analysis, caused by the superlinear growth of  the jump coefficient and the fact that higher moment bounds of the Poisson  increments contribute to magnitude not more than O(h). Numerical results are  finally reported to confirm the theoretical findings.