1.4k citations · 5.9k across the 68 of their papers we have counts for
4 papers · 2 filters
Error Preserving Correction for CPD and Bounded-Norm CPD
Anh-Huy Phan, Petr Tichavský, Andrzej Cichocki
In CANDECOMP/PARAFAC tensor decomposition, degeneracy often occurs in some difficult scenarios, e.g., when the rank exceeds the tensor dimension, or when the loading components are…
Best Rank-One Tensor Approximation and Parallel Update Algorithm for CPD
Anh-Huy Phan, Petr Tichavský, Andrzej Cichocki
A novel algorithm is proposed for CANDECOMP/PARAFAC tensor decomposition to exploit best rank-1 tensor approximation. Different from the existing algorithms, our algorithm updates…
Tensor Networks for Dimensionality Reduction and Large-Scale Optimizations. Part 2 Applications and Future Perspectives
A. Cichocki, A-H. Phan, Q. Zhao +4
Part 2 of this monograph builds on the introduction to tensor networks and their operations presented in Part 1. It focuses on tensor network models for super-compressed higher-ord…
Learning Efficient Tensor Representations with Ring Structure Networks
Qibin Zhao, Masashi Sugiyama, Andrzej Cichocki
Tensor train (TT) decomposition is a powerful representation for high-order tensors, which has been successfully applied to various machine learning tasks in recent years. However,…