In the RBF network, to overcome the defects of traditional K-means scheme with local search, an orthogonal least square algorithm is used to select RBF center.
在RBF网络中，为了克服传统K均值聚类法局部寻优的缺陷，采用了正交最小二乘法选取r BF中心。
The model USES an improved nearest-neighbor clustering algorithm to select the RBF center, and a recursive least square algorithm to train weights of the RBF neural network.
该模型首先采用改进的最近邻聚类算法确定径向基函数中心，接着应用递推最小二乘法训练网络的权值。
A recursive least squares algorithm for linking weight between linear layers of RBF network is introduced, and a dynamic recursive algorithm of center vector is proposed.
介绍了R BF网络线性层权值的训练算法——递推最小二乘法，及中心向量的动态递推算法。
应用推荐