Sparse and Truncated Nuclear Norm Based Tensor Completion
Research output: Journal Publications and Reviews (RGC: 21, 22, 62) › 21_Publication in refereed journal › peer-review
Author(s)
Related Research Unit(s)
Detail(s)
Original language | English |
---|---|
Pages (from-to) | 729-743 |
Number of pages | 14 |
Journal / Publication | Neural Processing Letters |
Volume | 45 |
Issue number | 3 |
Publication status | Published - Jun 2017 |
Link(s)
Abstract
One of the main difficulties in tensor completion is the calculation of the tensor rank. Recently a tensor nuclear norm, which is equal to the weighted sum of matrix nuclear norms of all unfoldings of the tensor, was proposed to address this issue. However, in this approach, all the singular values are minimized simultaneously. Hence the tensor rank may not be well approximated. In addition, many existing algorithms ignore the structural information of the tensor. This paper presents a tensor completion algorithm based on the proposed tensor truncated nuclear norm, which is superior to the traditional tensor nuclear norm. Furthermore, to maintain the structural information, a sparse regularization term, defined in the transform domain, is added into the objective function. Experimental results showed that our proposed algorithm outperforms several state-of-the-art tensor completion schemes.
Research Area(s)
- Multi-dimensional DCT, Tensor completion, Truncated nuclear norm
Citation Format(s)
Sparse and Truncated Nuclear Norm Based Tensor Completion. / Han, Zi-Fa; Leung, Chi-Sing; Huang, Long-Ting; So, Hing Cheung.
In: Neural Processing Letters, Vol. 45, No. 3, 06.2017, p. 729-743.Research output: Journal Publications and Reviews (RGC: 21, 22, 62) › 21_Publication in refereed journal › peer-review