报告题目：Adaptive multi-dimensional low-rank balance methods fort ensor completion problems
报告人：凌晨教授（杭州电子科技大学理学院）
时间：2021年5月29日（周六）上午10：00-11：00
地点：必赢76net线路唯一官方网站425报告厅
邀请人：白敏茹
摘要： This paper proposes two tensor product based approaches to tensor completion, which recovers missing entries of data represented by tensors. The proposed approaches are based on the tensor singular-value decomposition and the related tensor tubal rank, which are able to capture hidden information from tensors thanks to the balanced consideration of multi-dimensional low-rank features. Accordingly, new optimization formulations for tensor completion are proposed as well as two new algorithms for their solutions. Some computational results for color images, multi-spectral images, videos, and magnetic resonance imaging data recovery show that our approaches perform better than some existing state-of-the-art tensor-based completion methods.
专家简介：
凌晨，杭州电子科技大学理学院教授，博士生导师。现任中国运筹学会理事、中国运筹学会数学规划分会副理事长、中国经济数学与管理数学研究会副理事长、浙江省数学会常务理事、国际ESI期刊 Pacific Journal of Optimization编委、国际期刊Statistics，Optimization & Information Computing编委，国家自然科学基金委数理科学部评审专家。研究方向：非线性规划、变分不等式与互补问题、张量计算、多变量多项式优化、半无限规划、随机规划、多目标优化理论与应用等。近十年来，主持国家自科基金和浙江省自科基金各4项、其中省基金重点项目1项。在国内外重要刊物发表论文70余篇，其中SCI期刊论文50余篇，多篇发表在Math. Program.、SIAM J. on Optim.和 SIAM J.on Matrix Anal.and Appl.、COAP、JOTA、JOGO等。
