基于主成分聚类分析的不同地区贫困问题探究  

Research on Poverty in Different Areas Based on Principal Component and Cluster Analysis

作　　者：赵未名 

机构地区：[1]曲阜师范大学,山东 曲阜

出　　处：《统计学与应用》2021年第4期650-656,共7页Statistical and Application

摘　　要：2020年是我国脱贫攻坚的收官之年。在我国脱贫攻坚战取得了全面胜利后,脱贫与返贫交叉发生是我国目前面临的主要问题。因此,深入分析造成贫困的影响因素、对不同地区根据贫困程度进行聚类,从而有针对性找到脱贫方法是十分必要的。本文选取经济、人口结构、教育、医疗、文化、交通六个维度十一个指标,利用主成分分析进行降维处理,根据累积贡献率选取前三个主成分。随后进行系统聚类,利用Ward聚类方法将31个地区按贫困成因分为四类,并根据分类提出有效的建议。The year 2020 will be the end of China’s fight against poverty. After the victory of the battle against poverty in China, the intersection of poverty alleviation and return to poverty is the main problem faced by China. Therefore, it is very necessary to analyze the influencing factors of poverty and cluster different regions according to the poverty degree, so as to find targeted poverty alleviation methods. This paper selects eleven indicators from six dimensions, including economy, population structure, education, medical care, culture and transportation, and uses principal component analysis for dimensionality reduction. The first three principal components are selected according to the cumulative contribution rate. Then, Ward clustering method was used to divide the 31 regions into four categories according to the causes of poverty, and effective suggestions were put forward according to the classification.

关 键 词：主成分分析 系统聚类 贫困成因 

分 类 号：F32[经济管理—产业经济]