首页-->科学研究-->教学科研动态

IT讲坛2022年第17期-Low-rank optimization models for spectral compressed sensing

时间:2022-05-13 15:21:00 文章来源 :学科 浏览量:64

讲座题目:Low-rank optimization models for spectral compressed sensing

主讲人:杨在 教授

讲座时间:5月18日(星期三)下午15:00-16:30

#腾讯会议:945-934-560

个人简介:

照片5

杨在,西安交通大学数学与统计学院教授、博士生导师。20072009年分获中山大学应用数学专业本科和硕士学位,2014年获新加坡南洋理工大学电子电气工程专业博士学位。主要从事信息处理与无线通信的数学理论与方法研究,发表学术论文40余篇,被引用2700余次(谷歌学术)。任2017年欧洲信号处理会议Tutorial主讲人、IEEE高级会员和Signal Processing期刊编委。2019年获基金委优秀青年基金资助。

报告摘要:

Spectral compressed sensing refers to the recovery of a spectrally sparse signal from compressive measurements and has wide applications in radar and wireless communications. It is linked to low-rank matrix recovery by applying Kronecker and Carathéodory-Fejér theorems for Hankel and Toeplitz matrices, based on which convex and nonconvex optimization algorithms have been proposed. In this talk, we will introduce previous low-rank matrix recovery formulations and point out their limitations. After that, we present two new low-rank optimization models to resolve these limitations and demonstrate their effectiveness with convex and nonconvex algorithms.