径向基函数神经网络的再学习算法及其应用

Relearning Algorithm and Application of Redial Basis Function Neural Network

为了应用径向基函数神经网络逐步地识别待研究系统,文章针对径向基函数神经网络的再学习算法开展了深入的研究。应用严格的数学推理方法,将径向基函数神经网络的再学习问题转化为矩阵求逆的附加运算。详细给出了径向基函数神经网络再学习算法中增加新训练样本和增加新基函数的数学公式,同时对如何获取新的训练样本进行了研究。

In order to apply redial basis function neural network to recognize the given system step by step, this paper studies the problem of relearning algorithm of redial basis function neural network. Applies the strict mathematical reasoning, translates the problem of relearning algorithm of redial basis function neural network into the problem additional operation of matrix operation. Presents the mathematical expressions of adding new learning sample and adding new redial basis function of relearning algorithm, studies how to get the new learning sample.