首页 | 本学科首页   官方微博 | 高级检索  
     

融合自组织映射与Wang-Mendel方法的模糊规则提取
引用本文:於东军,谌贻华,于海瑛. 融合自组织映射与Wang-Mendel方法的模糊规则提取[J]. 南京理工大学学报(自然科学版), 2011, 35(6)
作者姓名:於东军  谌贻华  于海瑛
作者单位:南京理工大学计算机科学与技术学院,江苏南京,210094
基金项目:江苏省自然科学基金(BK2011371); 江苏高校优势学科建设工程; 南京理工大学自主科研专项计划(2011YDXM19)
摘    要:为解决经典Wang-Mendel方法提取模糊规则时得到的规则数目较多的问题,该文提出一个融合自组织映射(SOM)网络和Wang-Mendel方法的规则提取方案.利用SOM对原始样本进行学习,再使用Wang-Mendel方法从SOM的原型向量中提取规则.因为SOM的原型向量学习得到了原始样本的模式分布规律,并且原型向量的数量远小于原始学习样本的数量,所以该方法降低了提取的规则数目及耗费时间.有关混沌时间序列预测的实验结果表明了该文方法的有效性.

关 键 词:自组织映射  模糊规则提取  混沌时间序列  模式分布

Fuzzy Rule Extraction by Fusing SOM and Wang-Mendel Method
YU Dong-jun , CHEN Yi-hua , YU Hai-ying. Fuzzy Rule Extraction by Fusing SOM and Wang-Mendel Method[J]. Journal of Nanjing University of Science and Technology(Nature Science), 2011, 35(6)
Authors:YU Dong-jun    CHEN Yi-hua    YU Hai-ying
Affiliation:YU Dong-jun,CHEN Yi-hua,YU Hai-ying(School of Computer Science and Technology,NUST,Nanjing 210094,China)
Abstract:To solve the problem that too many fuzzy rules may be extracted by the classic Wang-Mendel method,a novel rule extraction method which combines self-organizing map(SOM)and Wang-Mendel method is presented.Original samples are firstly learned by the SOM,and rules are further extracted from the codebook vectors of the learned SOM by the Wang-Mendel method.The proposed method decreases the number of the extracted rules and the time consumption because the codebook vectors of the learned SOM achieves the pattern...
Keywords:self-organizing map  fuzzy rule extraction  chaotic time series  pattern distribution  
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号