基于非正交仿射离散小波变换的神经网络应用

Applications of neural network based on disorthogonal andaffine and discrete wavelet transforms

用非正交仿射离散小波变换时频局部化集中的优点与神经网络的有关理论 ,对前传神经网络的组织结构与简化提出了处理方案 。

By aid of the property of local concentration of time frequency domain and related theory of neural network in constructing a disorthogonal and affine and discrete wavelet transforms,a scheme is put forward to formulate the structure of neural network and to simplify operational procedure.Finally,this technique is applied in the approximation of wave function.