多民族地区农村贫困群体识别与扶贫效果研究-基于k-medios聚类与关联规则方法

文章基于多维贫困理论,对贵州省少数民族聚居D乡扶贫数据进行分析。首先通过多因子方法得到“家庭背景”、“民族与自身发展”、“地理环境”、“健康与劳动力”四个重要贫困维度,并基于k-medios聚类识别得到5个差异贫困群体。并基于关联规则,得到不同扶贫手段在差异群体中效果关系。分析发现（1）不同群体扶贫效果存在显著差异;（2）各群体中,“家庭背景维度”、“健康与劳动力维度”对扶贫效果影响最大;（3）教育扶贫在“弱势”及“一般”状态贫困群体中没有收到好效果,但对少数民族群体效果显著。相关建议包括:（1）针对少数民族占优群体,推广“教育+产业”扶贫模式;（2）采取预防干预,防止“良势”群体向“弱势”群体转化。

Study on the identification of rural poor groups and the effect of poverty alleviation in the multi-ethnic areas

Based on the multidimensional poverty theory, we analyzed the poverty alleviation data in D ethnic minorities Township in Guizhou Province. Firstly, four important poverty dimensions of "family background", "ethnicity and self-development", "geographical environment", and "health and labor force" are obtained through the multi-factor method, and five differential poverty groups are obtained based on k-medios cluster identification. Secondly, based on the association rule, the implementation of different poverty alleviation methods in different groups is associated with the ultimate poverty alleviation effect. The analysis found that (1) there are significant differences in the poverty alleviation effects of different groups; (2) among the groups, the "family background dimension" and "health and labor dimension" have the greatest impact on the poverty alleviation effect; (3) educational poverty alleviation is in the "weak" and "general " among the poor groups, but the effect is remarkable for minority groups. Suggestions: (1) Promote the "education + developing industries" poverty alleviation model for ethnic minority groups; (2) Take preventive interventions to prevent "good" groups from turning into "vulnerable" groups.