ONLINE BACKPROPAGATION IN 2-LAYERED NEURAL NETWORKS

被引：52

作者：

RIEGLER, P

BIEHL, M

机构：

[1] Inst. fur Theor. Phys., Julius-Maximilians-Univ., Wurzburg

来源：

JOURNAL OF PHYSICS A-MATHEMATICAL AND GENERAL | 1995年 / 28卷 / 20期

关键词：

D O I：

10.1088/0305-4470/28/20/002

中图分类号：

O4 [物理学];

学科分类号：

0702 ;

摘要：

We present an exact analysis of learning a rule by on-line gradient descent in a two-layered neural network with adjustable hidden-to-output weights (backpropagation of error). Results are compared with the training of networks having the same architecture but fixed weights in the second layer.

引用

页码：L507 / L513

页数：7

共 13 条

[1] A THEORY OF ADAPTIVE PATTERN CLASSIFIERS
AMARI, S
[J]. IEEE TRANSACTIONS ON ELECTRONIC COMPUTERS, 1967, EC16 (03): : 299 - +
[2] BACKPROPAGATION AND STOCHASTIC GRADIENT DESCENT METHOD
AMARI, S
[J]. NEUROCOMPUTING, 1993, 5 (4-5) : 185 - 196
[3] LEARNING BY ONLINE GRADIENT DESCENT
BIEHL, M
SCHWARZE, H
[J]. JOURNAL OF PHYSICS A-MATHEMATICAL AND GENERAL, 1995, 28 (03): : 643 - 656
[4] Chauvin Y., 1995, BACKPROPAGATION THEO
[5] ONLINE LEARNING IN THE COMMITTEE MACHINE
COPELLI, M
CATICHA, N
[J]. JOURNAL OF PHYSICS A-MATHEMATICAL AND GENERAL, 1995, 28 (06): : 1615 - 1625
[6] Hertz J., 1991, INTRO THEORY NEURAL
[7] HESKES T, 1993, MATH F NEURAL NETWOR
[8] OPPER M, IN PRESS PHYSICS NEU, V3
[9] RIEGLER P, UNPUB
[10] EXACT SOLUTION FOR ONLINE LEARNING IN MULTILAYER NEURAL NETWORKS
SAAD, D
SOLLA, SA
[J]. PHYSICAL REVIEW LETTERS, 1995, 74 (21) : 4337 - 4340

← 1 2 →