Timezone: »
Fully-Connected Tensor Network Decomposition
Yu-Bang Zheng · Ting-Zhu Huang · Xi-Le Zhao · Qibin Zhao · Tai-Xiang Jiang
The popular tensor train (TT) and tensor ring (TR) decompositions have achieved promising results in science and engineering. However, TT and TR decompositions only establish an operation between adjacent two factors and are highly sensitive to the permutation of tensor modes, leading to an inadequate and inflexible representation. In this paper, we propose a generalized tensor decomposition, which decomposes an $N$th-order tensor into a set of $N$th-order factors and establishes an operation between any two factors. Since it can be graphically interpreted as a fully-connected network, we named it fully-connected tensor network (FCTN) decomposition. The superiorities of the FCTN decomposition lie in the outstanding capability for characterizing adequately the intrinsic correlations between any two modes of tensors and the essential invariance for transposition. Furthermore, we employ the FCTN decomposition to one representative task, i.e., tensor completion, and develop an efficient solving algorithm based on proximal alternating minimization. Theoretically, we prove the convergence of the developed algorithm, i.e., the sequence obtained by it globally converges to a critical point. Experimental results substantiate that the proposed method compares favorably to the state-of-the-art methods based on other tensor decompositions.
Author Information
Yu-Bang Zheng (School of Mathematical Sciences, University of Electronic Science and Technology of China)
Ting-Zhu Huang (School of Mathematical Sciences, University of Electronic Science and Technology of China)
Xi-Le Zhao (School of Mathematical Sciences, University of Electronic Science and Technology of China)
Qibin Zhao (RIKEN AIP)
Tai-Xiang Jiang (School of Economic Information Engineering, Southwestern University of Finance and Economics)
More from the Same Authors
-
2021 : Bayesian Tensor Networks »
Kriton Konstantinidis · Yao Lei Xu · Qibin Zhao · Danilo Mandic -
2021 : Low-Rank Tensor Completion via Coupled Framelet Transform »
Jian-Li Wang · Ting-Zhu Huang · Xi-Le Zhao · Tai-Xiang Jiang · Michael Ng -
2021 : Bayesian Latent Factor Model for Higher-order Data: an Extended Abstract »
Zerui Tao · Xuyang ZHAO · Toshihisa Tanaka · Qibin Zhao -
2021 : Is Rank Minimization of the Essence to Learn Tensor Network Structure? »
Chao Li · Qibin Zhao -
2022 Poster: Tensor Wheel Decomposition and Its Tensor Completion Application »
Zhong-Cheng Wu · Ting-Zhu Huang · Liang-Jian Deng · Hong-Xia Dou · Deyu Meng -
2023 Poster: Transformed Low-Rank Parameterization Can Help Robust Generalization for Tensor Neural Networks »
Andong Wang · Chao Li · Mingyuan Bai · Zhong Jin · Guoxu Zhou · Qibin Zhao -
2023 Poster: Undirected Probabilistic Model for Tensor Decomposition »
Zerui Tao · Toshihisa Tanaka · Qibin Zhao -
2022 Poster: SPD domain-specific batch normalization to crack interpretable unsupervised domain adaptation in EEG »
Reinmar Kobler · Jun-ichiro Hirayama · Qibin Zhao · Motoaki Kawanabe -
2021 : Discussion Pannel »
Xiao-Yang Liu · Qibin Zhao · Chao Li · Guillaume Rabusseau -
2021 : Bayesian Tensor Networks »
Kriton Konstantinidis · Yao Lei Xu · Qibin Zhao · Danilo Mandic -
2021 Workshop: Second Workshop on Quantum Tensor Networks in Machine Learning »
Xiao-Yang Liu · Qibin Zhao · Ivan Oseledets · Yufei Ding · Guillaume Rabusseau · Jean Kossaifi · Khadijeh Najafi · Anwar Walid · Andrzej Cichocki · Masashi Sugiyama -
2020 Workshop: First Workshop on Quantum Tensor Networks in Machine Learning »
Xiao-Yang Liu · Qibin Zhao · Jacob Biamonte · Cesar F Caiafa · Paul Pu Liang · Nadav Cohen · Stefan Leichenauer -
2019 Poster: Deep Multimodal Multilinear Fusion with High-order Polynomial Pooling »
Ming Hou · Jiajia Tang · Jianhai Zhang · Wanzeng Kong · Qibin Zhao -
2011 Poster: A Multilinear Subspace Regression Method Using Orthogonal Tensors Decompositions »
Qibin Zhao · Cesar F Caiafa · Danilo Mandic · Liqing Zhang · Tonio Ball · Andreas Schulze-bonhage · Andrzej S CICHOCKI -
2011 Spotlight: A Multilinear Subspace Regression Method Using Orthogonal Tensors Decompositions »
Qibin Zhao · Cesar F Caiafa · Danilo Mandic · Liqing Zhang · Tonio Ball · Andreas Schulze-bonhage · Andrzej S CICHOCKI