农村居民人均消费的主成分分析与聚类分析.doc

农村居民人均消费的主成分分析与聚类分析 Principal component analysis and cluster analysis of rural residents per capita consumption 内容摘要：通过主成分分析给出我国2001、2006、2011年三年的农村居民人均消费及排名，对2011年农村居民人均消费地域数据进行模糊C均值聚类，将分为高级消费、中等消费、低级消费三类，与参考文献进行比较可知分类结果基本相同。模糊C均值聚类与主成分排名不一致，改进聚类方法，利用谱系聚类，选择离差平方和距离聚类，使得聚类结果和主成分排名相一致。并给出分类结果。根据排名结果，从多方面分析安徽省农村居民人均消费水平逐年提的原因。 关键词：主成分分析；；模糊C均值聚类；消费水平 Abstract: Through principal component analysis giving the rural residents per capita consumption region scores and rankings of 2001, 2006, 2011. Analysing rural residents in 2011 per capita consumption of geographical data by Fuzzy C means clustering. The region is divided into three class, the high consumption, moderate consumption, and lower consumption. Compared with the classification results of the references basically the same. The annual rankings without significant difference by Spearman rank correlation test.Fuzzy C-means clustering and principal component are in different rankings, To make clustering and principal component have consistent rankings, improved clustering methods by using hierarchical clustering and Select squared deviation and distance clustering. According to the effectiveness of hierarchical clustering, finding the best classification methods, Making Hierarchical clustering finger of the distance between the optimal number of clusters and the optimal class, dividing the area into 24 categories, giving the classification results. According to the ranking results, analyzing Reason for this improved every year The per capita consumption level of Anhui Province rural residents from various factors reasons. Key words: The principal component analysis; Spearman rank correlation test; Fuzzy C means 目 录 引言 1 1. 主成分分析 1 1.1主成分分析的基本思想 1 1.2主成分的数学模型 1 1.3主成分分析用于综合评价的一般步骤 2 1.4第一主成分得分及排名 4 1.5Spearman等级相关检验 5 2. 模糊C均值聚类 6 2.1模糊C值聚类 6 2.2模糊C值聚类的建模过程 6 2.3聚类结果及分析 7 3.谱系聚类 8 3.1谱系聚类的基本思想 8 3.2谱系聚类的步骤 9 3.3谱系聚类图 9 3.4地区分类结果 10 4. 安徽省农村居民人均消费水平分析 10 参考文献 13 附录1 14 附录2 17 引言 消费是经济增长的原动力。在人类社会的漫长历史中，正是由于人类