The Evolution of Direct Sampling: Towards Quantitative Identification from Qualitative Probes
数学专题报告
报告题目(Title)：The Evolution of Direct Sampling: Towards Quantitative Identification from Qualitative Probes
报告人(Speaker)：刘晓东 研究员 (中国科学院)
地点(Place)：教四 105
时间(Time)： 2025年 5 月 9 日（周五）15:00 --16:00
邀请人(Inviter)：李海刚
报告摘要
Sampling-type methods offer distinct advantages in inverse scattering: computational efficiency, implementation simplicity, measurement noise robustness, and reduced prior-information dependence. Nevertheless, conventional sampling techniques predominantly serve qualitative purposes, primarily designed for topological feature extraction such as coarse localization and geometric approximation of scatterers. This talk reveals the direct sampling method's quantitative identification capabilities. We rigorously demonstrate that the direct sampling method achieves both geometric characterization of unknown scatterers/sources and quantitative recovery of critical physical parameters—including precise reconstruction of source intensity distributions and identification of material constitutive properties.
主讲人简介
刘晓东，2004年本科毕业于北京师范大学，2010年在中科院数学院博士毕业后留所工作，先后任助理研究员，副研究员和研究员，期间于2011年到2013年在德国卡鲁理工大学做洪堡博士后。刘晓东主要研究方向是偏微分方程反问题，特别是基于波散射的数学理论和数值算法。目前，在SIAM系列，JCP, IP等杂志发表40余篇文章。基于这些工作，先后入选数学院陈景润未来之星计划，中科院青年创新促进会和国家高层次人才计划。
