报告题目:Linearized proximal algorithms for convex composite optimization and applications in sensor network localization
报告人：李冲教授
报告时间：15:30
报告地点：理A110
题目：Linearized proximal algorithms for convex composite optimization and applications in sensor network localization
时间：5月11日（周三）15:30
地点： 理A110
报告人：李冲教授
摘要： In the talk, we propose an efficient algorithm, called linearized proximal al-gorithm (LPA for short), for solving convex composite optimization problems, and study its convergence rates. The proposed LPA has the attractive computational advantage that the each subproblem is a strongly convex optimization problem, which avoids the di_culty as in the Gauss-Newton method (GNM) of _nding a solution with minimum norm among the set of minima of its subproblem. We show that it still maintains the same local convergence rate as that of the GNM under some mild assumptions. Also a globalization strategy for the LPA based on a backtracking line-search and an inexact version of the LPA are proposed and similar convergence results are established.
Applications in sensor network localization are provided, The sensor net-work localization problem is to determine the positions of the sensors in
a network by using the given incomplete pairwise distance measurements (truncated by the radio range). We formulate the sensor network localiza-tion
problem as a nonconvex feasibility problem, which can be cast into a convex composite optimization problem. We apply the LPA to solve a (pos-sibly nonconvex) feasibility problem, as well as a sensor network localization problem. Our numerical results illustrate that the LPA achieves the more
precise solution, costs less CPUtime and requires less information (the small radio range and the few anchors) than that of the semide_nite relaxation (SDR) technique, and thus, the LPA meets the demand for an e_cient and robust algorithm for the sensor network localization problem.
报告人简介：
李冲，浙江大学数学系教授、博士生导师。主要从事非线性逼近理论、非线性优化理论、科学计算等领域的研究。应邀多次访问香港中文大学、香港科技大学、台湾中山大学、澳大利亚新南威尔士大学、西班牙Sevillia大学、在国际顶级期刊Math. Program、SIAM J. Optim、SIAM J.Control Optim 、IMA J. Numer. Anal.和著名期刊J. London. Math. Soc.、Nonlinear Anal.、J. Approx. Theory、J. Complexity等上发表SCI学术论文150余篇，出版专著1部。主持承担完成国家自然科学基金面上项目10项、首届教育部新世纪优秀人才计划项目1项、省部级基金项目5项。
1992年被评为原商业部有突出贡献的中青年专家，享受政府特殊津贴。1995年获浙江省青少年英才奖。1998年入选为江苏省“青蓝工程”青年骨干教师。2002年获教育部优秀骨干教师。