Fast communication: Normalized natural gradient in independent component analysis

Authors:
Ying Tang;Jianping Li
Affiliations:
School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 610054, PR China;School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 610054, PR China
Venue:
Signal Processing
Year:
2010

Citing 16
Cited 0

How can we speed up matrix multiplication?

SIAM Review
An information-maximization approach to blind separation and blind deconvolution

Neural Computation
Stability analysis of learning algorithms for blind source separation

Neural Networks
Natural gradient works efficiently in learning

Neural Computation
An alternative perspective on adaptive independent component analysis algorithms

Neural Computation
Independent component analysis using an extended infomax algorithm for mixed subgaussian and supergaussian sources

Neural Computation
Contravariant adaptation on structured matrix spaces

Signal Processing
Adaptive Blind Signal and Image Processing: Learning Algorithms and Applications

Adaptive Blind Signal and Image Processing: Learning Algorithms and Applications
A novel adaptive learning rate sequential blind source separation algorithm

Signal Processing
The Scaling and Squaring Method for the Matrix Exponential Revisited

SIAM Journal on Matrix Analysis and Applications
Improving Generalization Performance of Natural Gradient Learning Using Optimized Regularization by NIC

Neural Computation
Nonholonomic Orthogonal Learning Algorithms for Blind Source Separation

Neural Computation
Adaptive Method of Realizing Natural Gradient Learning for Multilayer Perceptrons

Neural Computation
Fast communication: Stability analysis of natural gradient learning rules in overdetermined ICA

Signal Processing
A survey on methods for computing matrix exponentials in numerical schemes for ODEs

ICCS'03 Proceedings of the 2003 international conference on Computational science: PartII
Equivariant adaptive source separation

IEEE Transactions on Signal Processing

Quantified Score

Hi-index	0.08

Visualization

Abstract

This letter establishes the normalized natural gradient based algorithm in independent component analysis by employing simple matrix algebra. Moreover, another algorithm tightly related to the normalized natural gradient is introduced based on the truncated version of matrix exponentials, strictly restricting solutions in the space of non-singular matrices. Experimental results demonstrate the superior convergence of the proposed algorithms.