A stochastic and competitive network for the separation of sources

14 years 8 days ago

Download www.dice.ucl.ac.be

This paper presents an adaptive procedure for the linear and non-linear separation of signalswithnon-uniform,symmetricalprobabilitydistributions,basedonbothsimulatedannealing andcompetitivelearningmethodsbymeansofaneuralnetwork,consideringthepropertiesofthe vectorial spaces of signals, and using a multiple linearization in the mixture space.

Carlos García Puntonet, Ali Mansour, Manuel

Real-time Traffic

Adaptive Procedure | ESANN 2001 | ESANN 2007 | Non-linear Separation | Vectorial Spaces |

claim paper

» Joint Optimization of MAC and Network Coding for Cooperative and Competitive Wireless Mult...

» NeuralBased Separating Method for Nonlinear Mixtures

» Large scale manifold transduction

» Deep Learning Made Easier by Linear Transformations in Perceptrons

» Solving Independent Component Analysis Contrast Functions with Particle Swarm Optimization

» Switching Dynamics of Neural Systems in the Presence of Multiplicative Colored Noise

» BSS A New Approach for Watermark Attack

Post Info
More Details (n/a)

Added	31 Oct 2010
Updated	31 Oct 2010
Type	Conference
Year	2001
Where	ESANN
Authors	Carlos García Puntonet, Ali Mansour, Manuel Rodríguez Álvarez, Beatriz Prieto, Ignacio Rojas Ruiz

Comments (0)

Sciweavers

A stochastic and competitive network for the separation of sources

Adaptive Procedure | ESANN 2001 | ESANN 2007 | Non-linear Separation | Vectorial Spaces |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers