利用径向基函数改进算法实现函数逼近

Application of asymptotic function by RBFNN ameliorative arithmetic

分析了径向基函数神经网络(RBFNN)的原理,总结出径向基函数网络的一种实用插值算法.按照此方法对实际的例子进行了计算,结果表明,本算法快捷、可靠.

In order to solve the question of interpolating calibration , the principle of RBFNN is analyzed first for the purpose of better understanding, and then a useful kind of radius basis function neural networks （RBFNN） computation method is put forward. The method can be used to compute the data interpolation on line and surface. It is taken as another important networks other than BP in recent years. At last, an example is given to show this method is robust.