报告人:王川龙 教授 太原师范制服做爱
报告题目:Minimal and Maximal Truncated Nuclear Norm Regularization for Tensor Completion
时间:2025年7月18日(周五),下午17:00—18:00
地点:数学楼2-1会议室
摘要:
In this paper, we propose novel programming for tensor completion based on Tucker rank. The objective function of the problem is a weighted combination of the minimal and maximal truncated nuclear norm of N-mode matrices. The proximal gradient algorithm with extrapolation is proposed for solving the new programming. The objective function is proved to be the Kurdyka-Lojasiewicz function with an exponent of 1/2, which guarantees that the sequence produced by the algorithm converges globally to a stationary point. Finally, numerical results on some color images and video in painting problems show that the proposed optimization and algorithm has almost the same precision as the traditional several methods but with a significant reduction in CPU time.
报告人简介:
王川龙,男,1964年11月生,博士,教授,博士生导师,曾任中国工业与应用数学学会常务理事,现任山西省工业与应用数学学会理事长、全国运筹学会理事。山西省教学名师,山西省委联系高级专家, 山西省跨世纪学术和技术带头人,山西省青年学术带头人及“三晋英才”。2002年获山西省教学成果一等奖。2004年获山西省科技进步二等奖。2020、2024年分别获山西省自然科学奖二等奖。主持国家自然科学基金、山西省自然科学基金等项目10余项。著有《最优化原理与微观经济学》、《对角优势矩阵及其应用》等专著,在国内外学术期刊发表论文100余篇,其中被国际权威检索系统SCI收录60余篇。
邀请人:陈志平 教授