Chaotic dynamics in weight space of neural networks

被引：0

作者：

Gu, YQ ^{[1
]}

Huang, WQ ^{[1
]}

Chen, TL ^{[1
]}

机构：

[1] Nankai Univ, Dept Phys, Tianjin 300071, Peoples R China

来源：

COMMUNICATIONS IN THEORETICAL PHYSICS | 1999年 / 32卷 / 02期

关键词：

neural network; training algorithm; chaotic dynamics; nonlinear classification; forecasting nonlinear time series;

D O I：

暂无

中图分类号：

O4 [物理学];

学科分类号：

0702 ;

摘要：

When a special nonlinear self-feedback term is introduced into the dynamical equation of the backpropagation training algorithm for networks, the dynamics in weight space of networks can become chaotic. Chaotic dynamics of the system can help it escape from the most commonplace local minima of the energy. Simulation on the XOR problem and the prediction of chaotic time series have shown that the proposed chaotic training algorithm can converge to the global minimum or its approximate solutions efficiently and dramatically faster than the original backpropagation training algorithm.

引用

页码：247 / 252

页数：6

共 7 条

[1] Azoff E.M., 1994, NEURAL NETWORK TIME
[2] 1ST-ORDER AND 2ND-ORDER METHODS FOR LEARNING - BETWEEN STEEPEST DESCENT AND NEWTON METHOD
BATTITI, R
[J]. NEURAL COMPUTATION, 1992, 4 (02) : 141 - 166
[3] Nguyen D, 1990, P INT JOINT C NEUR N
[4] Rumelhart DE, 1986, PARALLEL DISTRIBUTED, V1, DOI DOI 10.7551/MITPRESS/5236.001.0001
[5] Werbos P. J., 1993, J INTELL FUZZY SYST, V1, P365
[6] 30 YEARS OF ADAPTIVE NEURAL NETWORKS - PERCEPTRON, MADALINE, AND BACKPROPAGATION
WIDROW, B
LEHR, MA
[J]. PROCEEDINGS OF THE IEEE, 1990, 78 (09) : 1415 - 1442
[7] Zhou CS, 1997, NEUROCOMPUTING, V14, P209, DOI 10.1016/S0925-2312(96)00030-6

← 1 →