Show simple item record

dc.contributor.advisor谭忠
dc.contributor.author侯晓宇
dc.date.accessioned2018-12-05T01:40:20Z
dc.date.available2018-12-05T01:40:20Z
dc.date.issued2017-12-27
dc.identifier.urihttps://dspace.xmu.edu.cn/handle/2288/170030
dc.description.abstract低秩矩阵填充问题(Low-rankMatrixCompletion)是指对于有部分位置上元素未知的矩阵,在假设矩阵低秩的前提下,可以通过优化算法来将其填充成一个完整的矩阵。低秩矩阵填充在机器学习、图像处理、推荐系统等领域发挥着重要的作用,是现今处理海量、高维数据的有力分析工具。本文首先介绍了低秩矩阵填充模型的理论发展,再分别根据将原模型进行凸松弛和非凸松弛后的改进模型综述了目前主要的算法,其中包括凸松弛的SVT算法,APG算法,ALM算法和非凸松弛的WMMN模型,并分析说明了不同算法在不同的领域,针对不同的模型有着各自的优势。 目前主要用于解决低秩矩阵填充的模型是用矩阵核范数来逼近目标函...
dc.description.abstractLow-rank matrix completion refers to problem that use optimization algorithm to fill a matrix which have unknown elements into a complete matrix.In general,we assume the incomplete matrix is a low-rank matrix.This methods have played an important role in areas such as machine learning, image processing, recommendation systems and so on,it's a powerful analysis tool of high-dimensional data.In this...
dc.language.isozh_CN
dc.relation.urihttps://catalog.xmu.edu.cn/opac/openlink.php?strText=58270&doctype=ALL&strSearchType=callno
dc.source.urihttps://etd.xmu.edu.cn/detail.asp?serial=60793
dc.subject低秩矩阵填充
dc.subjectL_1/2正则化
dc.subject加速近端梯度法
dc.subjectlow-rank matrix completion
dc.subjectL_1/2 regularization
dc.subjectAPG algorithm
dc.title基于S_1/2建模的低秩矩阵填充及其快速近端梯度算法的研究
dc.title.alternativeThe Low-rank Matrix Completion Based on S_1/2 Modeling and the Research of its APG Algorithm
dc.typethesis
dc.date.replied2017-05-26
dc.description.note学位:理学硕士
dc.description.note院系专业:数学科学学院_应用数学
dc.description.note学号:19020141152627


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record