应数学与计算科学学院及广西高校数据分析与计算重点实验室邀请，河南大学数学与统计学院肖运海教授、东莞理工学院信息科学学院程万友教授将于2021年5月14日上午来校讲学，欢迎全校师生踊跃参加。报告具体安排如下：

题目一：An Efficient Semismooth Newton Method for Adaptive Sparse Signal Recovery Problems

时间：2021年5月14日（周五）上午8：30

地点：花江校区6教6205室

主讲人：肖运海

摘要：In this talk, we propose an adaptive and robust model, say l_p-l_(1-2), where the l_p-norm with p≥1 measures the data fidelity and the l_(1-2)-term measures the sparsity. This proposed model is not only robust in the sense of having the ability to deal with different types of noises, but also can extract the sparse property even under high coherent condition.  However, minimizing this model is a challenging task because of the nonconvexity of the regularized term and the nonsmoothness of the data fidelity term. To overcome this difficulty, we use a proximal majorization-minimization technique to handle the nonconvex regularization term and then employ a semismooth Newton method to solve the corresponding convex relaxation subproblem. We prove that the sequence generated by the semismooth Newton method admits fast local convergence rate under some technical assumptions. Finally, we do some numerical experiments which demonstrate that the superiority of the proposed model is quite evident and the performance of the proposed algorithm is highly visible. This is a joint work with Y. Ding and H. Zhang from Beijing University of Technology, and P. Li from East China Normal University.

主讲人简介：

肖运海，博士，教授，博士生导师，河南省特聘教授。2007年博士毕业于湖南大学，曾在南京大学、台湾理论科学研究中心做博士后研究，曾经学术访问新加坡国立大学、香港理工大学、加拿大西蒙弗雷泽大学等。主要研究方向为稀疏优化、统计优化。在MPC、COAP、JSC、CSDA等发表论文40余篇，他引800余次，主持国家自然科学基金3项、参加973计划1项。目前担任中国工业与应用数学会理事、中国运筹学会数学规划分会理事、河南省运筹学会副理事长、河南大学学术委员会委员、河南大学应用数学研究所所长、《数学季刊》编辑部执行主任等。

题目二： A  Nonmonotone  Proximal Point  Algorithm for  Nonconvex  Regularized Optimization with Box Constraints

时间：2021年5月14日（周五）上午10：30

地点：花江校区6教6205室

主讲人：程万友

摘要：In this paper, we design a nonmonotone proximal point algorithm for a class of nonconvex sparsity-promoting penalties with box constraints, which include smoothly clipped absolute deviation (SCAD), capped L1 (CAP) and minimax concavity penalty (MCP). The new algorithm iteratively solves a proximal operator problem, which in turn utilizes a closed form solution of SCAD, CAP and MCP penalties with box constraints. To accelerate the convergence of the algorithm,  a nonmonotone line search strategy is used. We verify that any accumulation point of the sequence generated by the algorithm is a first-order stationary point . Furthermore, we  prove that  the worst-case iteration complexity for finding an scaled first-order stationary point is O（ε^-2）. The numerical experiments on various synthetic and real data demonstrate the efficiency of the proposed algorithm.

主讲人简介：

程万友， 博士，教授，中国数学规划分会青年理事、美国《Mathematical Reviews》的评论员。研究兴趣：最优化理论与算法及其在图像处理、信号处理及机器学习等问题中的应用。08年以来发表论文20多篇，其中被SCI期刊收录20多篇。现主持国家自然科学基金面项目1项、教育部人文社会科学研究一般项目和广东省自然科学基金自由申请项目各1项，主持完成国家自然科学基金青年基金和广东省自然科学基金面上项目各一项。 2014年入选广东省优秀青年教师培养对象, 2015年入选东莞市特色人才， 2016入选东莞理工学院高水平大学建设骨干。