| 基于改进kmeans聚类方法的RBF神经网络设计 |
| |
| 引用本文： | 何迎生,段明秀.基于改进kmeans聚类方法的RBF神经网络设计[J].邵阳学院学报(自然科学版),2008,5(2):48-50. |
| |
| 作者姓名： | 何迎生 段明秀 |
| |
| 作者单位： | 吉首大学,数学与计算机科学学院,湖南,吉首,416000 |
| |
| 摘 要： | 提出了一种利用改进的kmeans聚类算法确定径向基函数的数目、基函数中心及宽度，输出层权值由线性方程组确定的RBF网络设计方法．通过对函数逼近的仿真实验，表明该RBF网络比Kmeans聚类算法设计的RBF网络更加稳定高效．
|
| 关 键 词： | kmeam RBF 神经网络 中心 宽度 |
Design of RBF Network Based on Improved Kmeans Clustering |
| |
| HE Ying-sheng,DUAN Ming-xiu.Design of RBF Network Based on Improved Kmeans Clustering[J].Journal of Shaoyang University:Science and Technology,2008,5(2):48-50. |
| |
| Authors: | HE Ying-sheng DUAN Ming-xiu |
| |
| Institution: | (College of Mathematics and Computer Science,Jishou University, Jishou, Hunan 416000) |
| |
| Abstract: | In the paper ,we propose a method of design RBF Neural Network using the improved Kmeans Clustering algorithm to determine the number and center and spread of Redial Basis Function.The simulate experiment of function approximation explains the RBF Neural Network more stable and high efficiency than the RBF Neural Network designing by the Kmeans clustering algorithm. |
| |
| Keywords: | kmeans RBF neural network center spread |
| 本文献已被 维普 万方数据 等数据库收录！ |
