论文部分内容阅读
将小波变换原理与非线性时间序列逼近问题相结合 ,利用模型分析方法 ,对一阶非线性自回归过程 (NLAR(1) )的小波神经网络逼近问题进行了研究 ,构造出了逼近函数的具体表达式 ,并给出了NLAR(1)过程所逼近的误差限
Combining the principle of wavelet transform with the approximation of nonlinear time series, the approximation of wavelet neural networks for first-order nonlinear autoregressive (NLAR (1)) is studied by using the model analysis method, and the concrete approximation function is constructed Expression, and gives the error limit approached by the NLAR (1) process