报告时间:2024年11月15日(星期五)上午9:30-10:30
报告地点:理学楼609报告厅
报告题目:Machine learning methods for scattering and inverse scattering problems
报告摘要: In this talk, I will describe the latest progress we have made in solving scattering problems and inverse scattering problems. For the unbounded domain scattering problem, we have designed an alternatively-optimized SNN, which is fast and accurate. Meanwhile, we design a machine learning strategy that can solve the multi-body scattering problem without prior knowledge of the number of obstacle bodies.
报告人简介:吕俊良,吉林大学6163银河线路检测中心教授,博士生导师。研究兴趣包括无界域散射问题的数值方法,反散射问题的数值算法与理论,偏微分方程有限体积元法等。研究成果发表在 SIAM J. Numer. Anal.,Math. Comput.,Inverse Problems, J. Sci. Comput. 以及 IMA J. Numer. Anal. 等杂志上。承担国家自然科学基金项目、国防项目、吉林省自然科学基金等。现任中国数学会计算数学分会理事、中国仿真学会仿真算法专委会委员、吉林省数学会理事、美国数学会数学评论员。