MLP+H: a hybrid neural architecture formed by the interaction of hopfield and multi-layer perceptron neural networks

Authors:
Clayton Silva Oliveira;Emílio Del Moral Hernandez
Affiliations:
Polytechnic School of the University of São Paulo, Department of Eletronic Systems Engineering, Cidade Universitária, Siio Paulo, SP, Brazil;Polytechnic School of the University of São Paulo, Department of Eletronic Systems Engineering, Cidade Universitária, Siio Paulo, SP, Brazil
Venue:
IWANN'03 Proceedings of the Artificial and natural neural networks 7th international conference on Computational methods in neural modeling - Volume 1
Year:
2003

Citing 10
Cited 0

Blind separation of sources, Part 1: an adaptive algorithm based on neuromimetic architecture

Signal Processing
Independent component analysis, a new concept?

Signal Processing - Special issue on higher order statistics
An information-maximization approach to blind separation and blind deconvolution

Neural Computation
A fast fixed-point algorithm for independent component analysis

Neural Computation
Natural gradient works efficiently in learning

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

Neural Computation
Semiparametric model and superefficiency in blind deconvolution

Signal Processing
A blind source separation technique using second-order statistics

IEEE Transactions on Signal Processing
Equivariant adaptive source separation

IEEE Transactions on Signal Processing
Blind source separation-semiparametric statistical approach

IEEE Transactions on Signal Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we study the blind source separation problem of temporally correlated signals via exploring both the temporal structure and high-order statistics of source signals. First, we formulate the problem as independent residual analysis and present a simple cost function. Efficient learning algorithm is developed for the demixing matrix and the corresponding stability analysis is also provided. The formulation provides much more flexibility for us to identify learning algorithms with good learning performance and stability. Furthermore, the approach unifies the conventional high-order statistical method and the second-order statistical method. From stability analysis, we infer that if the temporal filters of sources are mutually different, the second order statistical algorithm will be sufficient to separate the sources from their linear mixtures.