基于BP神经网络的湖南省县域贫困识别研究  

Research on Poverty Recognition of County in Hunan Province Based on BP Neural Network

作　　者：袁帅[1] 刘雨昕 汪意 刘一诺 李鹏[1] YUAN Shuai;LIU Yu-xin;WANG Yi;LIU Yi-nuo;LI Peng(School of Geographical Sciences,Hunan Normal University,Changsha,410081,China)

机构地区：[1]湖南师范大学地理科学学院,长沙410081

出　　处：《湖南师范大学自然科学学报》2022年第6期32-40,共9页Journal of Natural Science of Hunan Normal University

基　　金：湖南省大学生创新创业训练基金项目(2019010,S201910542010)。

摘　　要：本文以湖南省为研究区域,借助2017年社会经济数据包括居民收入、城市化水平、国内生产总值等,通过BP人工神经网络模拟贫困结果。结果表明:贫困测度标准需考虑多维因素影响,主要包括城市化水平、人均GDP、人均收入支出及第三产业生产总值;夜间灯光遥感可以反映地区社会经济发展,对贫困分布有参考价值;贫困水平较高以上的县域有63个,多分布在湘西、株洲南部、邵阳北部;湖南省的贫困县分布具体为西部县市(27个)、中部县市(10个)、东部县市(9个);湖南省主要致贫类型归因于社会发展水平,武陵山、罗霄山连片贫困区基础设施和社会建设较落后、人才资金流失导致两极分化严重。对贫困分布和致贫因素的分析,有利于准确认识贫困成因,为脱贫验收和巩固成效提供参考。Based on socioeconomic data of 2017,including resident income,the level of urbanization and GDP,back-propagation(BP) artificial neural network was used to simulate the poverty of Hunan province in this research.Our results showed that the poverty measurement standard needs to consider the influence of multi-dimensional factors,mainly including the level of urbanization,GDP per capita,income per capita,and GDP of the tertiary industry.The remote sensing of night lights can reflect the social and economic development of the region and have a reference value for the distribution of poverty.There are 63 counties in higher poverty level in Hunan province,mainly distributing in Xiangxi,southern zhuzhou and northern shaoyang.Of which the western,the central and the eastern counties are 27,10 and 9,respectively.The main poverty type in Hunan province is attributed to social development level.The infrastructure and socia construction up the adjacent poor areas in Wuling Mountain and Luoxiao Mountain are backward,and the serious loss of talent and capital leads to the polarization.The analysis of the distribution of poverty and the factors that cause poverty is conducive to the accurate understanding of the causes of poverty and to provide reference for the effect of poverty alleviation and consolidation.

关 键 词：贫困测度 BP神经网络 县域经济 夜间灯光遥感 

分 类 号：C912.81[经济管理]