摘要
模糊聚类分析是模糊模式识别中一个重要研究领域,而其中最经典的模糊C均值算法认为样本矢量各特征对聚类结果贡献均匀,没有考虑不同的属性特征对模式分类的不同影响,在处理属性高相关的数据集时,该算法分错率增加。针对这些问题,提出了一种基于马氏距离特征加权的模糊聚类算法,利用自适应马氏距离的优点对特征加权处理,对高属性相关的数据集进行更有效的分类。实验证明该方法的可行性和有效性。
Fuzzy clustering analysis is an important research field of the fuzzy pattern recognition,and the Fuzzy C-Means algorithm（FCM）is the most classical algorithm.It regards the sample features have the same contribution to the cluster result;not thinking the different features may have different impacts on the cluster result.When FCM processes some datasets of high correlation,error probability will be increased.Focusing on above two problems,this paper proposes an improved new fuzzy clustering algorithm based on feature weighted Mahalanobis distance function.Using adaptive Mahalanobis distance to weight the feature,the new algorithm can effectively cluster to the datasets of high correlation.Experiment illustrates its effectiveness and feasibility.
出处
《计算机工程与应用》
CSCD
2012年第5期198-200,共3页
Computer Engineering and Applications
基金
国家自然科学基金(No.10771092)
辽宁省科技厅博士启动基金(No.20081079)
辽宁省教育厅高等学校科研项目资助(No.2008347)
关键词
模糊C均值
马氏距离
属性相关
特征加权
Fuzzy C-Means
Mahalanobis distances
correlation of attributes
feature weighted
