Title :
Competitive learning neural networks applied to multivariate data set reduction
Author :
Delsert, Stephane ; Hamad, Denis ; Daoudi, Mohamed ; Postaire, Jack-Gerard
Author_Institution :
Centre d´´Automatique de Lille, Lille I Univ., Villeneuve d´´Ascq, France
Abstract :
Presents three competitive learning neural networks applied to the multivariate data set reduction problem. The synaptic vectors of a neural network are used as prototypes of the data set. The quality of the results are compared, using an example, by means of an informational criterion. This criterion evaluates the quality of the matching between the density function estimated from the whole data set and that determined from the reduced set
Keywords :
data reduction; neural nets; unsupervised learning; competitive learning neural networks; data set prototypes; density function; informational criterion; matching quality; multivariate data set reduction; synaptic vectors; Density functional theory; Image analysis; Image processing; Image storage; Neural networks; Probability density function; Prototypes; Speech analysis; Speech processing; Vector quantization;
Conference_Titel :
Systems, Man and Cybernetics, 1993. 'Systems Engineering in the Service of Humans', Conference Proceedings., International Conference on
Conference_Location :
Le Touquet
Print_ISBN :
0-7803-0911-1
DOI :
10.1109/ICSMC.1993.390762
