基于支持向量机的模糊小波神经网络  

Fuzzy Wavelet Neural Networks Based on SVM

作　　者：张晓光[1] 匡颖芝[1] 耿道华 吴行标[1] 

机构地区：[1]中国矿业大学机电工程学院 [2]江苏天能集团陈楼煤矿,徐州221000

出　　处：《华东理工大学学报（自然科学版）》2006年第11期1351-1354,1368,共5页Journal of East China University of Science and Technology

基　　金：江苏省博士后科研基金资助项目(0502010B);中国矿业大学科技基金资助项目(2005B005)

摘　　要：针对高维输入小波网络的初始参数和网络结构非常复杂且计算量大的问题,提出用支持向量机(SVM)确定小波网络的初始参数和网络结构的方法。首先,使用有监督模糊聚类算法从聚类中抽取模糊规则,然后对每一个规则的后件使用支持向量机方法确定小波网络的结构和初始参数,最后采用梯度下降方法调节模糊小波网络中的参数,使得模糊小波网络输出与期望输出之间的误差较小。仿真结果表明:该算法与传统的模糊神经网络(FNN)相比显著提高了分类精度。For wavelets network with high-dimensional inputs, both the initial parameters and network structure are very complicated, and the amount of computation is large. Toward this problem, a new method is proposed using support vector machine （SVM） to ascertain both the initial parameters and network structure of wavelet network. At first, the supervised fuzzy clustering algorithm is used to extract fuzzy rule from clustering. And then, for the consequent rule, SVM is applied to both the initial parameters and the network structure of wavelet network. Finally, gradient descent algorithm is adopted to adjust the parameters of fuzzy wavelet network so as to make the error between the outputs of fuzzy wavelet network and the expectant outputs is less. Simulation result shows that this algorithm can greatly improve the classification precision comparing to traditional fuzzy neural network（FNN）.

关 键 词：模糊小波神经网络 支持向量机 有监督模糊聚类 分类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]