基于FCM的两级集成分类器算法  

Classifier algorithm of two-level ensemble based on FCM

作　　者：齐淼[1] 张化祥[1] 赵蕾 

机构地区：[1]山东师范大学信息科学与工程学院,山东济南250014 [2]山东省平阴县环境卫生处,山东济南250400

出　　处：《计算机工程与设计》2010年第16期3658-3661,3665,共5页Computer Engineering and Design

基　　金：山东省中青年科学家科研奖励基金项目(2006BS01020);山东省高新技术自主创新工程专项计划基金项目(2007ZZ17);山东省自然科学基金项目(Y2007G16);山东省科技攻关计划基金项目(2008GG10001015);山东省教育厅科技计划基金项目(J07YJ04);山东省信息产业发展专项基金项目(2008B0026)

摘　　要：基于模糊聚类的思想提出了一种新的两级集成分类器算法。将数据集用Fuzzy C-Means算法进行聚类,得到每个实例对应于每个类别的模糊隶属度。一级集成根据Bagging算法获得成员分类器,分类器个数为数据集类别数且每个成员分类器对应一个类别标号,这些成员分类器的采样方式是通过其对应类别的模糊隶属度为每个实例加权后进行随机重采样。二级集成是将一级集成产生的针对类别的成员分类器通过动态加权多数投票法来组合,学习到最终的分类结果。该算法称为EWFuzzyBagging,实验结果表明,该算法与Bagging和AdaBoost相比具有更好的健壮性。A novel two level ensemble classifier algorithm based on fuzzy clustering is proposed.First,fuzzy c-means method is used to cluster the instances,and then every instance obtains the fuzzy membership corresponding to its class label.The first level ensemble obtains component classifiers by bagging algorithm.The number of classifiers is equal to the number of class labels of datasets,while every component classifier corresponding to one class label.The sampling way of these component classifiers is random resample by means of weighting every instance of training set with the membership degree of the corresponding class.The second level ensemble is used to combine the component classifiers aiming at classes,which produced by the first level,through dynamic weighted majority voting,final classify result is learned.The new approach is called EWFuzzyBagging,which shows more robust comparing to Bagging and AdaBoost.

关 键 词：模糊聚类 模糊隶属度 BAGGING 动态加权 分类器集成 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]