The function approximation ability comparison of two wavelet networks and their applications
Qian Qing-quan · Relay · 2003
In this paper, from two different aspects of wavelet transformations in wavelet analysis, two wavelet networks are induced, and the constructions,models and function approximation abilities of the two wavelet networks are presented. The simulations of a ladder function approximation with the two wavelet networks are done, and the result is that the function approximation ability of the second wavelet network is better than the first. The two wavelet networks used as signal classification are discussed, and the learning models and simulation results are given in the end.