混合神经模糊分类器的实现  

Implementation of Hybrid Neuro- fuzzy Classifier

作　　者：刘淑英[1] 

机构地区：[1]咸阳师范学院信息工程学院,陕西咸阳712000

出　　处：《计算机技术与发展》2013年第12期113-115,共3页Computer Technology and Development

基　　金：国家自然科学基金资助项目(40572082);陕西省教育科研计划项目(11JK1046)

摘　　要：人工神经网络与模糊系统是计算智能的核心内容,二者的混合系统是近年来的一个研究热点。分类是数据分析中的研究重点,随着数据的复杂化和多样化,对分类的要求越来越高,有时仅凭经验和专业知识难以确切地进行分类,因此研究如何运用神经模糊分类算法进行数据分析具有重要意义与实用价值。鉴于其强大的数据分析功能,研究中采用模糊C均值聚类算法和Gath-Geva聚类算法对数据进行分类,并对测试数据进行仿真试验,其测试结果良好。Artificial neural network and fuzzy system were considered the main components of computation intelligence, the hybrid system about them was one of study topics in recent years. Classification is a research focus in data analysis, as data is complicated and diversi- fied, the requirements for classification will be increasingly high, sometimes only by experience and professional knowledge not to accu- rately classify. In view of their powerful data analysis functions, using neuro-fuzzy algorithm for data analysis will be meaningful and useful. In this paper,fuzzy C-means clustering algorithm model and Gath-Geva clustering algorithm model are proposed for the parame- ter classification, which is simulated, and obtain good results.

关 键 词：神经网络 模糊系统 聚类 模糊C均值 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]