文件名称:xiaobosjinwangl
介绍说明--下载内容均来自于网络,请自行研究使用
利用多分辨分析方法,结合小波分析和神经网络思想构建一种新型的神经网络模型———小波神经网络,解决了传
统神经网络中隐层节点数难以确定的问题。通过对股票的预测,说明该方法能有效地提高预测精度, 避免了人工神经网
络模型的固有缺陷。
-Using multi-resolution analysis method, combined with wavelet analysis and neural network ideological construct a new neural network model--- wavelet neural network to solve the traditional neural network hidden layer nodes is difficult to determine. Forecast by the stock, indicating that the method can effectively improve the prediction accuracy of artificial neural network model to avoid the inherent shortcomings.
统神经网络中隐层节点数难以确定的问题。通过对股票的预测,说明该方法能有效地提高预测精度, 避免了人工神经网
络模型的固有缺陷。
-Using multi-resolution analysis method, combined with wavelet analysis and neural network ideological construct a new neural network model--- wavelet neural network to solve the traditional neural network hidden layer nodes is difficult to determine. Forecast by the stock, indicating that the method can effectively improve the prediction accuracy of artificial neural network model to avoid the inherent shortcomings.
(系统自动生成,下载前可以参看下载内容)
下载文件列表
基于小波神经网络的时间序列流数据的研究.pdf