Application on Machine Learning Method in Accurate Identification of Poverty Alleviation Objects—Taking the Survey Data of Residents in Certain Western Mountain Area as an Example

Abstract

The government has an important duty, that is, reducing the social poverty, so how to recognition poverty alleviation objects becomes a critical factor. We selected the province western mountain area for the location of sampling, selecting index designed questionnaire to conduct on-the-spot investigation, collection of mountainous area residents family specific data, using the method of multidimensional poverty measure UNDP-MPI ideas of poverty indicators for dimensions, using the analytic hierarchy process (AHP) to empowerment of multidimensional index, multidimensional poverty measurement model is established. The multidimensional poverty measurement model is used to identify the poor families. Finally, the multidimensional poverty measurement model is fitted and verified with the single income index model, so as to determine the effectiveness of the multidimensional poverty measurement model and solve the problem of poor households' inefficient identification.