Identification of Poor Households in Precision Poverty Alleviation Based on Ensemble Learning

Abstract

Poverty alleviation has always been a major problem that plagues national economic development and people's livelihood. Through the research on precise poverty alleviation, it is hoped to find a feasible way, its operating mechanism and principle, so as to improve the effect of poverty alleviation. The purpose of this paper is to study the identification of poor households in precision poverty alleviation based on ensemble learning. This paper introduces the current research status of precise poverty alleviation and the application of ensemble learning algorithms in various fields, and discusses some advantages of boosting and XGBoost in classification, paving the way for the following. Combined with the actual situation of M County, the algorithm index system has been expanded to better reflect the poverty status of farmers. The ensemble learning method is applied to the poverty identification problem, and the model evaluation standard is used to measure the effectiveness and stability of multiple models. The experimental results show that the XGBoost model in this paper has the best application effect in the identification of poor households, with an accuracy rate of 98.2%.