聚类分析在我国城镇居民消费结构研究中的应用  

Application of Clustering Analysis to Research Consumption Structure of Households in Chinese Suburb Area

作　　者：张月[1] 张金凤[1] 王泓娜[1] 关静[1] 

机构地区：[1]辽宁大学数学系,辽宁沈阳110036

出　　处：《辽宁大学学报（自然科学版）》2004年第1期47-49,共3页Journal of Liaoning University：Natural Sciences Edition

摘　　要：选用了8个聚类分析统计指标用离差平方和法,对各地区城镇居民消费结构相似程度进行了聚类分析,不仅从总体上掌握了消费结构类型的地区分布,而且系统阐述了各地区消费结构的特点及产生原因,同时提出用恩格尔系数来衡量生活水平高低的局限性,得出在确定地区生活水平评价体系时应加入非食品性支出构成的使用力度的结论.In this paper, eight statistics indexes of clustering analysis are chosen regarding consumptive structure of households in Chinese suburb area. Clustering analysis is applied in consumption structure similarity degree of households by means of deviation squared summation. Not only we master the region distribution of consumptive structure type generally, but also introduce the character and reason of each type. At the same time we point out the limitation of Engel coefficient to measure the level of living. We draw a conclusion that we should strengthen the use of non - food expenditure composition when we determine judgement index system of region level of living.

关 键 词：聚类分析 消费结构 相似程度 恩格尔系数 离差平方和法 

分 类 号：F126.1[经济管理—世界经济] F224.7