Paper Info
Title
An Adaptive Nonunitary Joint Diagonalization Algorithm of High-Order Tensors.
Abstract
Joint diagonalization (JD)of high-order tensors is a generalization of an approximate JD algorithm of a series of target matrices. Considering the number of target tensors may increase with time, we present an adaptive nonunitary JD algorithm of high-order tensors. The algorithm recursively minimizes a least squares criterion and updates diagonalizer matrices by the online algorithm and tensor calculations. The complexity of our algorithm is lower than the iterative algorithms. Simulation results demonstrate that the proposed algorithm is efficient for JD of high-order tensors.
Year
DOI
Venue
2018
10.1109/CISP-BMEI.2018.8633078
CISP-BMEI
Keywords
Field
DocType
Signal processing algorithms,Approximation algorithms,Complexity theory,Machine learning algorithms,Cost function,Computational modeling
Least squares,Approximation algorithm,Online algorithm,Tensor,Computer science,Matrix (mathematics),Algorithm,Recursion,Signal processing algorithms
Conference
ISBN
Citations 
PageRank 
978-1-5386-7604-2
0
0.34
References 
Authors
0
3
Name
Order
Citations
PageRank
Jing Xia16611.85
Guanghui Cheng221.40
Jifei Miao301.69