Chaotic Dynamics in Weight Space of Neural Networks |
| |
Authors: | GU YuQiao HUANG WuQun CHEN TianLun |
| |
Institution: | Department of Physics, Nankai University, Tianjin 300071, China |
| |
Abstract: | 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. |
| |
Keywords: | neural network training algorithm chaotic dynamics nonlinear classification forecasting nonlinear time series |
|
| 点击此处可从《理论物理通讯》浏览原始摘要信息 |
| 点击此处可从《理论物理通讯》下载免费的PDF全文 |