Prediction Model of the Poverty Return Risk from the Perspective of Poverty Alleviation
According to the 2021 poverty alleviation account data provided by Kaili Rural Revitaliza-tion Bureau,a statistical model for risk measurement is established.First,the mixed polynomial model is used to build the annual income prediction model.Secondly,the logistics regression model is used to build the poverty risk prediction model,and the linear classification model of"three categories of households"is obtained by combining SVM and other machine learning algorithms.Then,the non par-ametric regression model with the annual per capita annual income is built through the evaluation score data.Through the analysis of multiple models,it provides technical assistance for grass-roots units to carry out poverty alleviation risk screening.