|本期目录/Table of Contents|

[1]豆 蔻,吴云韬*,黄龙庭,等.基于张量链分解的低秩张量补全研究[J].武汉工程大学学报,2021,43(04):442-447.[doi:10.19843/j.cnki.CN42-1779/TQ.202104010]
 DOU Kou,WU Yuntao*,HUANG Longting,et al.Low-Rank Tensor Completion Based on Tensor Train Decomposition[J].Journal of Wuhan Institute of Technology,2021,43(04):442-447.[doi:10.19843/j.cnki.CN42-1779/TQ.202104010]
点击复制

基于张量链分解的低秩张量补全研究(/HTML)
分享到:

《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]

卷:
43
期数:
2021年04期
页码:
442-447
栏目:
机电与信息工程
出版日期:
2021-08-31

文章信息/Info

Title:
Low-Rank Tensor Completion Based on Tensor Train Decomposition
文章编号:
1674 - 2869(2021)04 - 0442 - 06
作者:
豆 蔻1吴云韬*1黄龙庭2陈 里3
1. 武汉工程大学计算机科学与工程学院,湖北 武汉 430205; 2. 武汉理工大学信息学院,湖北 武汉 430070;3. 武汉工程大学邮电与信息工程学院,湖北 武汉 430074
Author(s):
DOU Kou1 WU Yuntao*1 HUANG Longting2 CHEN Li3
1. School of Computer Science and Engineering, Wuhan Institute of Technology, Wuhan 430205, China;2. School of Information Engineering, Wuhan University of Technology, Wuhan 430070, China;3. The College of Post and Telecommunication of WIT, Wuhan 430074, China
关键词:
张量补全张量链分解核范数张量分解
Keywords:
tensor completion tensor train decomposition nuclear norm tensor decomposition
分类号:
TN911.72
DOI:
10.19843/j.cnki.CN42-1779/TQ.202104010
文献标志码:
A
摘要:
为了进一步提高低秩张量补全性能,针对基于传统张量分解方法的张量补全问题研究中的计算复杂问题,根据张量链分解能够将高阶张量分解成一组三阶核心张量进行有效降维的特点,本文基于张量链分解的核心张量模型,采用核范数最小化方法求解,对缺失张量的低秩补全问题进行了研究,并且分别在实际图像以及合成数据上进行了算法对比实验,实验结果证明了本文方法的有效性,与目前流行的方法相比,运行速度更快、收敛性更好、补全结果也较优。
Abstract:
To further improve the performance of low-rank tensor completion, for the current computational complexity in the study of tensor completion based on traditional tensor decomposition methods, high-order tensors can be decomposed into a set of third-order core tensors according to tensor train decomposition. Based on the core tensor model of tensor train decomposition, this paper uses the nuclear norm minimization method to study the problem of low-rank completion of missing tensors, and algorithm comparison experiments were carried out respectively based on images and synthetic data. The experimental results prove the effectiveness of the proposed method, and compared with the current popular methods, it runs faster, has better convergence and completion results.

参考文献/References:

[1] HUANG L T, ALMEIDA A L F, SO H. Target estimation in bistatic MIMO radar via tensor completion[J]. Signal Processing, 2016, 120: 654-659. [2] HAN K, NEHORAI A. Nested vector-sensor array processing via tensor modeling[J]. IEEE Transactions on Signal Processing, 2014, 62(10): 2542-2553. [3] 贾慧迪, 韩志, 陈希爱, 等. 基于非局部张量火车分解的彩色图像修补[J]. 模式识别与人工智能, 2019, 32(10): 955-963. [4] 柳欣, 钟必能, 张茂胜, 等. 基于张量低秩恢复和块稀疏表示的运动显著性目标提取[J]. 计算机辅助设计与图形学学报, 2014, 26(10): 1753-1763. [5] CONG F, LIN Q H, KUANG L D, et al. Tensor decomposition of EEG signals: a brief review[J]. Journal of Neuroscience Methods, 2015, 248: 59-69. [6] 杨兵. 基于张量数据的机器学习方法研究与应用[D]. 北京:中国农业大学, 2014. [7] BRO R. PARAFAC. Tutorial and applications[J]. Chemometrics Intelligent Laboratory Systems, 1997, 38(2): 149-171. [8] TUCKER L R. Some mathematical notes on three-mode factor analysis[J]. Psychometrika, 1966, 31(3): 279-311. [9] ACAR E, DUNLAVY D M, KOLDA T G, et al. Scalable tensor factorizations for incomplete data[J]. Chemometrics Intelligent Laboratory Systems, 2011, 106(1): 41-56. [10] ZHAO Q B, ZHANG L, CICHOCKI A. Bayesian CP factorization of incomplete tensors with automatic rank determination[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2015, 37(9): 1751-1763. [11] CHEN Y L, HSU C T, LIAO H Y M. Simultaneous tensor decomposition and completion using factor priors[J]. IEEE transactions on Pattern Analysis Machine Intelligence, 2013, 36(3): 577-591. [12] OSELEDETS I V. Tensor-train decomposition[J]. SIAM Journal on Scientific Computing, 2011, 33(5): 2295-2317. [13] 刘园园. 快速低秩矩阵与张量恢复的算法研究[D]. 西安:西安电子科技大学, 2013. [14] BOYD S, PARIKH N, CHU E. Distributed optimization and statistical learning via the alternating direction method of multipliers[M]. Delft: Now Publishers Inc, 2011. [15] KOLDA T G, BADER B W. Tensor decompositions and applications[J]. SIAM Review, 2009, 51(3): 455-500. [16] CAI J F, CANDèS E J, SHEN Z. A singular value thresholding algorithm for matrix completion[J]. SIAM Journal on Optimization, 2010, 20(4): 1956-1982. [17] LIU J, MUSIALSKI P, WONKA P, et al. Tensor completion for estimating missing values in visual data[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2012, 35(1): 208-220. [18] YUAN L, ZHAO Q, CAO J. Completion of high order tensor data with missing entries via tensor-train decomposition[C]// International Conference on Neural Information Processing. Cham: Springer, 2017: 222-229. [19] YUAN L,ZHAO Q,GUI L,et al. High-order tensor completion via gradient-based optimization under tensor train format[J]. Signal Processing: Image Communication, 2019, 73: 53-61.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2021-04-07基金项目:国家自然科学基金(61771353)作者简介:豆 蔻,硕士研究生。E-mail:zllx0951@qq.com *通讯作者:吴云韬,博士,教授。E-mail:ytwu@wit.edu.cn引文格式:豆蔻,吴云韬,黄龙庭,等. 基于张量链分解的低秩张量补全研究[J]. 武汉工程大学学报,2021,43(4):442-447.
更新日期/Last Update: 2021-08-07