报告题目:Iterative hard thresholding alogrithm and its variations

报告人：沈益 教授（浙江理工大学）

报告时间：2024.04.09 16:00－17:00

报告地点：理A110

报告摘要：The Iterative hard thresholding algorithm and its variations are widely used in many applications including image processing, one-bit compressed sensing, sparse principal component analysis and so on. Numerical experiments show that some modified algorithms have good performance, while theoretical guarantees are not provided in the literatures. In this talk, we mainly study the convergence properties of these algorithms. The mathematical models can be viewed as a general continuously cost function subject to sparsity constraints.

报告人简介：沈益，浙江理工大学教授，博导，浙江省应用数学研究会副理事长；毕业于浙江大学数学系，获应用数学博士学位；从事应用调和分析、信息论、逼近论等相关领域的研究；主持国家自然科学基金面上项目、优秀青年科学基金项目、浙江省杰出青年基金项目等省部级项目；在Appl Comput Harmon A、IEEE T Inform Theory、IEEE T Signal Proces、J Approx Theory、J Fourier Anal Appl、Comput Aided Geom D、J Complexity等期刊发表SCI论文20余篇，曾获“浙江省优秀数学教师”称号。

邀请人：金永阳、沈守枫