基于特征加权的混合型模糊聚类分析及其在洪水灾害等级划分中的应用  

Mixed fuzzy clustering analysis based on feature weights and its application in classification of flood disaster grade

作　　者：孔令燕[1] 夏乐天[1] 

机构地区：[1]河海大学理学院,江苏南京210098

出　　处：《水利经济》2009年第5期1-3,共3页Journal of Economics of Water Resources

基　　金：国家自然科学基金(10671032);河海大学自然科学基金(2084408319)

摘　　要：提出了基于特征加权的混合型模糊聚类分析方法,即在改进的ISODATA方法的基础上,运用Relief F算法确定样本特征的权重。将这种分析方法应用到新疆"96.7"洪水灾害等级划分中,结果表明该方法是可行的。A mixed fuzzy clustering analysis method based on feature weights was put forward.In other words,the weights of sample characteristics were obtained by use of the relief F method based on an improved ISODATA method.This method was applied to the classification of Xinjiang '96.7' flood disaster grade. The results show that the present method is of good applicability.

关 键 词：特征加权 模糊聚类 RELIEF F算法 洪水灾害 

分 类 号：P426.616[天文地球—大气科学及气象学]