学 术

分享到微信 ×
打开微信“扫一扫”
即可将网页分享至朋友圈
博士后学术沙龙(第18期)
文:唐小青 来源:党委教师工作部、人力资源部(教师发展中心) 时间:2017-11-23 5667

  为搭建我校博士后之间的学术交流平台,促进学术水平提升,学校博士后管理办公室组织开展博士后学术沙龙活动。本次沙龙由我校博士后陈林、陈刚和王成祥分享其研究成果,诚挚邀请感兴趣的师生参加。

  一、时 间:2017年11月27日(周一)14:30

  二、地 点:清水河校区经管楼宾诺咖啡

  三、主办单位:电子科技大学博士后管理办公室

  四、承办单位:数学科学学院

         电子科技大学博士后联谊会

  五、活动安排:

  报告一:     

  (1)主题:Sample Average Approximation Method Based on Regularized Gap Function for Stochastic Mixed Variational Inequality Problems

  (2)主讲人: 陈林  数学科学学院博士后 

  (3)交流内容:

  In this talk, the sample average approximation method for the Expected-value formulation of stochastic mixed variational inequality problems (SMVIP) is proposed. The existence and uniqueness of solution of the Expected-value formulation of SMVIP are investigated under some suitable coercivity condition and uniformly monotonicity. Besides, a constrained optimization reformulation is proposed by using a regularized gap function. Under some mild conditions, an implementable sample average approximation method for this reformulation is established and the limiting behaviors of the optimal values and the optimal solutions of the approximation problems as the sample size increases are investigated as well. Some numerical results are also obtained to show the effectiveness of the proposed method.

  主讲人简介:

  L. Chen received the Bachelor degree from Chongqing Normal University and Ph.D. degree from Sichuan University, in 2009 and 2017, respectively. He currently works as a Postdoctoral Fellow in the School of Mathematical Sciences of UESTC. Dr. Chen’s primary researches are the theoretical and practical aspects of optimization. His recent interest is in optimal decision making problems that involve uncertain data. This includes developing and/or investigating new mathematical models which capture the uncertainty and other features such as competition, equilibrium and hierarchical relationships between decision makers. 

  报告二:

  (1)主题:Analysis of hybridizable discontinuous Galerkin finite element method for time-harmonic Maxwell equations

  (2)主讲人: 陈刚  数学科学学院博士后 

  (3)交流内容:

  We consider the time-harmonic Maxwell equations with smooth coefficients and Dirichlet boundary condition. The formulation of the Maxwell equations we consider here is a mixed curl-curl formulation, where the divergence-free condition is imposed by introducing a Lagrange multiplier. In the first part, we prove regularities of Maxwell equations explicit dependent on the frequency. In the second part, we introduce and analyze a hybridizable discontinuous Galerkin (HDG) method for the time-harmonic Maxwell equations. All the unknowns, as long as the unknowns at inter-element faces, of the underlying system are approximated by discontinuous polynomials. This method is shown be be stable and optimal convergent with respect to the mesh size, disrespect to the frequency, under minimal regularity. In the third part, we give a convergence analysis on the proposed HDG method for Maxwell eigenvalue. Numerical experiments with both smooth and singular analytical solutions are presented.    

  (4)主讲人简介:

  G. Chen received the Bachelor degree and Ph.D. degree from Sichuan University in 2014 and 2017, respectively. He currently works as a Postdoctoral Fellow in the School of Mathematics Sciences of UESTC. Dr. Chen is working in the field of finite element methods (FEMs). His main research interests include HDG FEMs for PDEs with parameter perturbation; stabilized FEMs for PDEs; FEMs for optimal control problems; HDG FEMs for eigenvalue problems.

  报告三:

  (1)主题:An adaptive iteration reconstruction method for limited-angle CT image reconstruction

  (2)主讲人:王成祥  数学与科学学院博士后 

  (3)交流内容:

  The limited-angle computed tomography (CT) reconstruction problem is an ill-posed inverse problem, and the parameter selection for limited-angle CT iteration reconstruction is a difficult issue in practical application. In this paper, to alleviate the instability of limited-angle CT reconstruction problem and automatize the reconstruction process, we propose an adaptive iteration reconstruction method that the regularization parameter is chosen adaptively via the plot of the normalized wavelet coefficients fitting residual versus that the L0 regularization part. The experimental results show that the reconstructed image using the method with adapted regularization parameter are almost as good as that using the non-adapted parameter method in terms of visual inspection, in addition, our method has an advantage in adaptively choosing the regularization parameter.

  主讲人简介:

  Chengxiang Wang received the Bachelor degree from Sichuan University of Arts and Science and Ph.D. degree from the Chongqing University, in 2011 and 2016, respectively. From 2012 to 2016, he did doctoral research at Engineering Research Center of Industrial Computed Tomography Nondestructive Testing of the Education Ministry of China, Chongqing, China. He currently works as a Postdoctoral Fellow in the School of Mathematical Sciences of UESTC. His research interest includes CT reconstruction, imaging method, image processing and optimization method.

 

                   电子科技大学博士后管理办公室

                      2017年11月22日


编辑:罗莎  / 审核:林坤  / 发布:林坤

"