关键词: 农村商业医疗保险, k近邻算法, 决策树算法, 逻辑回归算法, 集成学习, 潜在客户识别

Abstract: In order to promote the development of rural commercial medical insurance, the k-DT-LR fusion model based on k-nearest neighbor algorithm, decision tree algorithm and logical regression algorithm was proposed to dynamically allocate effective capabilities to each learner in the integration. Using CGSS2017 household survey data, the potential customer identification model of rural commercial medical insurance was constructed. The experimental results showed that the classification accuracy of k-DT-LR algorithm was 90.024% and the recall rate was 91.402%, which could accurately identify the potential customers of rural commercial medical insurance.

Key words: rural commercial medical insurance, k-nearest neighbor algorithm, decision tree algorithm, logistic regression algorithm, integrated learning, potential customer identification