理学论坛第一百三十七次学术活动(刘歆+张在坤报告)
发布时间: 2018-06-14   浏览次数: 13

报告题目1A Parallelizable Algorithm for Orthogonally Constrained Optimization Problems

报告人1:刘歆 副研究员 国家优秀青年基金获得者http://lsec.cc.ac.cn/~liuxin/

报告人1单位:中国科学院数学与系统科学研究院

时间20186201430-1530

地点:仙林校区教2-314

主办单位:理学院、视觉认知计算与应用研究中心、科研院

报告内容1To construct a parallel approach for solving orthogonally constrained optimization problems is usually regarded as an extremely difficult mission, due to the low scalability of orthogonalization procedure. In this talk, we propose an infeasible algorithm for solving optimization problems with orthogonality constraints, in which orthogonalization is no longer needed at each iteration, and hence the algorithm can be parallelized. We also establish a global subsequence convergence and a worst-case complexity for our proposed algorithm. Numerical experiments illustrate that the new algorithm attains a good performance and a high scalability in solving discretized Kohn-Sham total energy minimization problems.

报告人1简介:2004年本科毕业于北京大学数学科学学院;2009年于中国科学院研究生院获得博士学位,导师是袁亚湘院士。毕业后留所工作至今。期间分别在德国ZIB研究所、美国RICE大学、美国纽约大学Courant研究所进行过长期访问。主要研究方向包括正交约束矩阵优化问题,包括线性与非线性特征值问题;非线性最小二乘问题的算法与理论;分布式优化算法设计。刘歆主持并完成一项国家自然科学基金青年基金项目;现主持一项国家自然科学基金面上项目,并于20168月获得国家自然科学基金委优秀青年科学基金。201412月入选中国科学院数学中国运筹学会青年科技奖;20172月入选中国科学院北京分院“启明星”优秀人才计划。于20157月起担任《Mathematical Programming Computation》编委,于20177月起担任《计算数学》编委。



报告题目2A Continuous Optimization Model for Clustering

报告人2Dr. Zaikun ZhangEmailzaikun.zhang@polyu.edu.hk)(http://mat.uc.pt/~zhang/

报告人2单位The Hong Kong Polytechnic University

时间20186201530-1630

地点:仙林校区教2-314

主办单位:理学院、视觉认知计算与应用研究中心、科研院

报告内容2We study the problem of clustering a set of objects into groups according to a certain measure of similarity among the objects. This is one of the basic problems in data processing with various applications ranging from computer science to social analysis. We propose a new continuous model for this problem, the idea being to seek a balance between maximizing the number of clusters and minimizing the similarity among the objects from distinct clusters. Adopting the methodology of spectral clustering, our model quantifies the number of clusters via the rank of a graph Laplacian, and then relaxes rank minimization to trace minimization with orthogonal constraints. We analyze the properties of our model, propose a block coordinate descent algorithm for it, and establish the global convergence of the algorithm. We then demonstrate our model and algorithm by several numerical examples.

This is a joint work with Xin Liu (Chinese Academy of Sciences), Michael Ng (Hong Kong Baptist University), and Rui Zhang (Chinese Academy of Sciences).

报告人2简介:张在坤博士,香港理工大学应用数学系助理教授。张在坤2007年本科毕业于吉林大学数学学院,2012年博士毕业于中科院数学与系统科学研究院,导师是袁亚湘院士。他毕业后访问科英布拉大学、欧洲科学计算研究与高等培训中心和图卢兹大学,从事博士后研究,于2016年加入香港理工大学应用数学系。张在坤的研究领域为最优化算法与理论,主要研究兴趣是无导数方法、随机化方法、子空间方法和带有噪声的优化问题,目前主持主持香港研究资助局ECS项目一项以及PROCORE---香港/法国合作研究项目一项(与图卢兹大学及空客公司IRTSaintExupéry研究所合作)