Title :
Comparison of Lagrange constrained neural network with traditional ICA methods
Author :
Szu, Harold ; Kopriva, Ivica
Author_Institution :
Digital Media RF Lab., George Washington Univ., DC, USA
fDate :
6/24/1905 12:00:00 AM
Abstract :
The paper presents comparison between the a priori MaxEnt and the a posteriori MaxEnt methodologies, namely Lagrange Constraint Neural Network (LCNN) by Szu in 1997 and ICA algorithms by Bell-Sejnowski-Amari-Oja (BSAO) and many others since 1996. We chose the remote sensing application because it is the only real world application that we know to be truly linear, single path, and instantaneous mixing of the unknown ground spectral objects
Keywords :
constraint handling; entropy; neural nets; transfer functions; LCNN; Lagrange Constraint Neural Network; a posteriori MaxEnt; a priori MaxEnt; mixing matrices; transfer function; Covariance matrix; Data models; Entropy; Filtering; Independent component analysis; Lagrangian functions; Neural networks; Remote sensing; Stochastic processes; Vectors;
Conference_Titel :
Neural Networks, 2002. IJCNN '02. Proceedings of the 2002 International Joint Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
0-7803-7278-6
DOI :
10.1109/IJCNN.2002.1005517
