学术报告一
报告题目: Perturbation theory of metric regularity
报告人: 何诣然教授
报告时间:2018年5月19日(星期六)上午9:00--9:30
报告地点:数学与统计学院报告厅(25教14楼)
参加人员:本科生、研究生、教师
摘要: It is shown, in the metric space setting, that the p-order metric regularity is preserved under perturbation by locally holder continuous mappings of order 1/p, extending a classical result which says that 1-order metric regularity is preserved under perturbation by locally Lipschitz mappings. The argument is based on a new result of fixed point theorem type.
报告人简介:何诣然,男,教授,博士生导师,四川师范大学科研处处长。2001获香港中文大学博士学位,2002年至2004年在新加坡国立大学做博士后。四川省学术与技术带头人后备人选,四川省有突出贡献的优秀专家。2009年入选中组部“西部之光”计划,全国先进工作者,全国五一劳动奖章获得者。研究领域为最优化理论,在SIAM J. Optim., Math. Oper. Res., J. Global Optim.等期刊发表论文27篇,他引100余次。
学术报告二
报告题目: Random Relaxed Projection Algorithms for Convex Minimization
报告人: 夏福全教授
报告时间:2018年5月19日(星期六)上午9:40--10:10
报告地点:数学与统计学院报告厅(25教14楼)
参加人员:本科生、研究生、教师
摘要: In this talk, we deal with relaxed projection algorithm with random feasibility steps for solving constrained convex minimization problems, where the constrained setis specified as the intersection of possibly infinitely many constraint sets ,and the objective function is the sum of a large number of component functions. Each constraint setis assumed to be a level set of a convex but not necessarily differentiable function. The relaxed projection algorithm is considered for randomization schemes and for cyclic schemes for the component functions and the constraint sets. Under some suitable assumptions, the method is shown to be convergent to the solution of convex minimization problem in almost sure sense. Preliminary computational experience is also reported.
报告人简介:夏福全,男,教授,博士生导师。2008年获四川大学博士学位。多次应邀访问韩国Gyeongsang大学,Gyungnam大学;台湾高雄医学大学基础科学教育中心,台湾长庚大学管理学院。 从事最优化和变分不等式理论及应用方面的研究,到目前为止,已在 J. Optim. Theory Appl., J. Global Optim.,Math. Methods Oper. Res. 等期刊上表论文30余篇,他引90余次。
学术报告三
报告题目: Existence and boundedness of solutions to maximal monotone inclusion problem
报告人: 张永乐博士
报告时间:2018年5月19日(星期六)上午10:30--11:00
报告地点:数学与统计学院报告厅(25教14楼)
参加人员:本科生、研究生、教师
摘要:In Hilbert spaces, the inclusion problem with an arbitrary maximal monotone operator is considered. We prove that the non-emptiness of the solution set of the inclusion problem is equivalent to a coercivity condition. Moreover, a sufficient and necessary condition for the boundedness of the solution set is obtained.
报告人简介:张永乐,男,博士,讲师。2012年博士毕业于四川师范大学数学与软件科学学院,同年留校工作。研究领域为最优化理论。在J. Optim. Theory Appl., Nonlinear Anal. 等期刊发表论文6篇。
学术报告四
报告题目:Optimal Control Problems in Large Time
报告人: 张灿博士
报告时间:2018年5月19日(星期六)上午11:10--11:40
报告地点:数学与统计学院报告厅(25教14楼)
参加人员:本科生、研究生、教师
摘要:In this talk, we will introduce periodic and exponential turnpike properties for the optimal control problems. The approach to analyze these turnpike theorems is from the view point of Pontryagin maximum principle as well as the saddle point theory of Hamiltonian system.
报告人简介: 张灿,男,博士,讲师。2014年博士毕业于武汉大学数学与统计学院,先后在武汉大学,法国巴黎第六大学,西班牙巴斯克大学做博士后,2017年被聘为武汉大学讲师。在 J. Math. Pures Appl., J. Eur. Math. Soc.,Ann. Inst. H. Poincaré Anal. Non Linéaire, SIAM J. Control Optim. 等期刊发表论文12篇。