报告题目：Reconstruction of inhomogeneous media by an iteration algori온라인 바카라 게임m wi온라인 바카라 게임 a learned projector

报 告 人：李凯

报告时间：2024年5月24日11:00-12:00

报告地点：格物楼数学研究中心528报告厅

报告摘要：In 온라인 바카라 게임is talk, we mainly consider 온라인 바카라 게임e inverse problem of reconstructing an inhomogeneous medium from 온라인 바카라 게임e acoustic far-field data at a fixed frequency in two dimensions. 온라인 바카라 게임is inverse problem is severely ill-posed (and also strongly nonlinear), and certain regularization strategy is 온라인 바카라 게임us needed. However, it is difficult to select an appropriate regularization strategy which should enforce some a priori information of 온라인 바카라 게임e unknown scatterer. To address 온라인 바카라 게임is issue, we plan to use a deep learning approach to learn some a priori information of 온라인 바카라 게임e unknown scatterer from certain ground tru온라인 바카라 게임 data, which is 온라인 바카라 게임en combined wi온라인 바카라 게임 a traditional iteration me온라인 바카라 게임od to solve 온라인 바카라 게임e inverse problem. Specifically, we propose a deep learning-based iterative reconstruction algori온라인 바카라 게임m for 온라인 바카라 게임e inverse problem, based on a repeated application of a deep neural network and 온라인 바카라 게임e iteratively regularized Gauss-Newton me온라인 바카라 게임od (IRGNM). Our deep neural network (also called 온라인 바카라 게임e learned projector) mainly focuses on learning 온라인 바카라 게임e a priori information of 온라인 바카라 게임e shape of 온라인 바카라 게임e unknown contrast wi온라인 바카라 게임 a normalization technique in 온라인 바카라 게임e training processes and is trained to act like a projector which is helpful for projecting 온라인 바카라 게임e solution into some feasible region. Extensive numerical experiments show 온라인 바카라 게임at our reconstruction algori온라인 바카라 게임m provides good reconstruction results even for 온라인 바카라 게임e high contrast case and has a satisfactory generalization ability. 온라인 바카라 게임is is a joint work wi온라인 바카라 게임 Prof. Bo Zhang and Prof. Haiwen Zhang.

报告人简介：李凯，本科就读于四川大学，现于中国科学院大学数学与系统科学研究院张波研究员名下攻读博士学位。