Title :
Training linear neural network with early stopped learning and ridge estimation
Author :
Vigneron, Vincent ; Barret, Claude
Author_Institution :
CEMIF, Evry Univ., France
Abstract :
This paper addresses the problem of supervised learning in layered neural network with linear units and includes an analysis of the effect of noise on training algorithms. We survey most of the known results on linear networks. The connections to classical statistical ideas such as ordinary least squares are emphasized
Keywords :
feedforward neural nets; learning (artificial intelligence); least squares approximations; principal component analysis; sensitivity analysis; least squares regression; linear neural network; multilayer neural network; principal component analysis; ridge estimation; sensitivity analysis; supervised learning; Algorithm design and analysis; Artificial neural networks; Collaboration; Computer networks; Covariance matrix; Indexing; Least squares methods; Neural networks; Supervised learning; Vectors;
Conference_Titel :
Neural Networks, 1998. Proceedings. Vth Brazilian Symposium on
Conference_Location :
Belo Horizonte
Print_ISBN :
0-8186-8629-4
DOI :
10.1109/SBRN.1998.730987
