报告题目:Smoothing strategies for support vector regression and signal reconstruction
报告人:陈界山(台湾师范大学)
报告时间:2022年7月6日 15:00~16:00
报告地点:腾讯会议 会议号:554 976 5868
报告内容简介:In this talk, we present two smoothing strategies to support vector regression and signal reconstruction problem, which are popular target problems in data science. We demonstrate how smoothing functions can be involved or constructed to fit in these two types of problems. For support vector regression, we employ smoothing Newton algorithm to work along with the proposed smoothing functions, whereas the conjugate algorithm is adopted to tackle with the signal reconstruction problem.
报告人简介:陈界山,1967年,台湾师范大学特聘教授,理学院院长。2004年毕业于美国华盛顿大学数学系,获博士学位(师从国际著名优化大师Paul Tseng教授)。主要从事连续优化理论与应用、非光滑分析等研究,尤其在二阶锥优化理论与应用领域做出了一系列的杰出工作。在Mathematical Programming,Mathematics of Operations Research,SIAM Journal on Optimization等国际学术期刊发表130余篇论文,独立发表学术专著“SOC Functions and Their Applications(2019年,Springer出版社)”、“Complementarity Functions in Optimization(2022年,Springer出版社)”两部。目前担任国际SCI期刊Taiwanese Journal of Mathematics副主编、Pacific Journal of Optimization编委。个人主页:http://math.ntnu.edu.tw/~jschen/index.php?menu=Home
邀请人:矫立国 (助理研究员 jiaolg356@nenu.edu.cn)