第37期
题目:Inexact Optimization Algorithms Using Fenchel-Rockafellar Duality
主讲人:王浩,研究员
上海科技大学信息科学与技术学院
时间:2019年7月11日上午9:00
地点:海韵行政楼C510
报告摘要:The primary focus of this work is to design algorithms for solving large-scale problems with each subproblem being solved inexactly. The performance of this type of algorithms crucially replies on the termination criterion of terminating the subproblem solve. This tutorial propose a new termination criterion using Fenchel-Rockafellar Duality. Based on this technique, three inexact algorithms were designed. The first one is an inexact sequential quadratic programming algorithm (SQP), the second one is an inexact successive linear programming (SLP), and the last one is an inexact gradient projection method. By controlling the inexactness of the subproblem solution, we can significantly reduce the computational cost needed at each iteration. Global convergence for both feasible and infeasible cases is proved. Complexity analysis for the optimality residual is also derived under loose assumptions. Numerical experiments exhibit the ability of the proposed algorithm to rapidly find optimal solution through cheap computational cost.
报告人简介:王浩博士于2015年4月在美国理海大学工业工程系获得博士学位,并于2010年和2007年在北京航空航天大学数学与应用数学系分别获得理学硕士和学士学位。在攻读博士期间,曾于2012年、2014年和2015年分别在埃克森美孚企业战略实验室、三菱电机研究实验室和群邑集团研发部担任实习研究员。当前研究领域主要为非线性优化、大规模优化、稀疏优化以及分布式优化。主要成果在优化顶级期刊SIAM Journal on Optimization上发表。
第38期
题目:Generalized Sparse and Low-Rank Optimization for Ultra-Dense Networks
主讲人:石远明,研究员
上海科技大学信息科学与技术学院
时间:2019年7月11日 上午10:30
地点:海韵行政楼C510
报告摘要:Ultra-dense network (UDN) is a promising technology to further evolve wireless networks and meet the diverse performance requirements of 5G networks. With abundant access points, each with communication, computation and storage resources, UDN brings unprecedented benefits, including significant improvement in network spectral efficiency and energy efficiency, greatly reduced latency to enable intelligent mobile applications, and the capability of providing massive access for Internet of Things (IoT) devices. However, such great promises come with formidable research challenges. To design and operate such complex networks with various types of resources, efficient and innovative methodologies will be needed. This motivates the recent introduction of highly structured and generalizable models for network optimization. This tutorial shall present recent advances in structured sparse and generalized low-rank techniques for optimizing UDNs, with a comprehensive coverage including modeling, algorithm design, and theoretical analysis. Through motivating applications (e.g., mobile edge caching and wireless distributed learning), the powerfulness of this set of tools will be demonstrated, and their abilities in solving key design problems in UDNs will be highlighted. A special attention is paid on algorithmic approaches to deal with nonconvex objective functions and constraints, as well as computational scalability.
报告人简介:石远明博士于2011年7月获得清华大学电子工程学士学位(2007-2009学年于“数学物理基础科学班”培养);2015年8月获得香港科技大学电子及计算机工程博士学位,师从Khaled B. Letaief教授。他于2015年9月加入上海科技大学信息科学与技术学院任助理教授/研究员,2019年1月任上海科技大学信息科学与技术学院常任副教授/研究员。他于2016年秋季学期任加州大学伯克利分校访问教授,访问统计机器学习领域Martin J. Wainwright教授。他的研究成果荣获2016 IEEE通信学会马可尼最佳论文奖(无线通信领域最重要学术奖项之一),以及2016 IEEE信号处理学会最佳青年作者论文奖。他主要的研究方向为运筹优化、高维统计、机器学习、信号处理,及其在无线通信、量化金融中的应用。
邀请人:信息与通信工程系 付立群教授