Tensor completion using total variation and low-rank matrix factorization

详细信息    查看全文

• 作者：Teng-Yu Ji ^{tengyu_j66@126.com" class="auth_mail" title="E-mail the corresponding author} ; Ting-Zhu Huang ; ^{tingzhuhuang@126.com" class="auth_mail" title="E-mail the corresponding author} ; ^{tzhuang@uestc.edu.cn" class="auth_mail" title="E-mail the corresponding author} ; Xi-Le Zhao ^{xlzhao122003@163.com" class="auth_mail" title="E-mail the corresponding author} ; Tian-Hui Ma ^{nkmth0307@126.com" class="auth_mail" title="E-mail the corresponding author} ; Gang Liu ^{wd5577@163.com" class="auth_mail" title="E-mail the corresponding author}
• 关键词：Tensor completion ; Total variation ; Low-rank matrix factorization ; Block coordinate descent
• 刊名：Information Sciences
• 出版年：2016
• 出版时间：1 January 2016
• 年：2016
• 卷：326
• 期：Complete
• 页码：243-257
• 全文大小：2594 K

文摘

In this paper, we study the problem of recovering a tensor with missing data. We propose a new model combining the total variation regularization and low-rank matrix factorization. A block coordinate decent (BCD) algorithm is developed to efficiently solve the proposed optimization model. We theoretically show that under some mild conditions, the algorithm converges to the coordinatewise minimizers. Experimental results are reported to demonstrate the effectiveness of the proposed model and the efficiency of the numerical scheme.