摘要
在有色噪声环境下，通常的神经网络不能进行有效学习．本文引入一个噪声处理器对神经网络加以改进，使之能进行有效学习，并在运行阶段能跟踪系统输出．一个仿真例子用来说明改进后神经网络的性能．
In color noise environment, general neural network can not learn efficiently.In this paper, a noise processor is introduced to improve the neural network's property. So it can learn efficiently and can trace the output of the system in running time. A simulated example is given to represent the property of the improved neural network.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
1998年第5期68-71,共4页
Journal of Southeast University：Natural Science Edition
关键词
有色噪声
神经网络
结构
color noise
neural network
structure
