摘要
心脑血管疾病是老龄人群患病率、死亡率最高的疾病之一,带来了沉重的经济与社会负担。随着我国人口老龄化程度加剧,开发针对带病体的商业健康保险产品显得尤为重要。本文基于详实的历史住院数据,采用机器学习方法对心脑血管疾病人群的住院费用进行预测,对不同机器学习方法进行预测比较后,采用表现最佳的LightGBM对心脑血管疾病群体进行个人自付费用的测算,分析了性别、年龄、省份、社保类型、婚姻状况、合并症和地区经济条件对个人自付费用的影响,并在此基础上预测了保险理赔情况,发现“年龄”是影响理赔的主要因素。本研究结果可为保险公司设计创新型商业健康保险产品提供借鉴。
The number of patients with cardiovascular and cerebrovascular disease in China has exceeded 330 million.These diseases pose a grave health risk to the elderly,representing one of the highest morbidity and mortality rates amongst the aging population.As the population aging intensifies in China,the development of related commercial health insurance products becomes increasingly pertinent.This paper,drawing on comprehensive historical hospitalization data,employs machine learning techniques to forecast hospitalization expenses for individuals with cardiovascular and cerebrovascular conditions.It also compares the predictive capabilities of various machine learning methodologies.After the comparison,the LightGBM,which demonstrates superior performance,is used to estimate out-of-pocket expenses for individuals within this disease group.It also studies the impacts on personal out-of-pocket expenses by factors such as gender,age,province,type of social insurance,marital status,comorbidities,and regional economic conditions.It then further predicts on insurance claims,identifying′age′as a primary factor affecting claims.Insurance companies can leverage on the results of this research to design innovative commercial health insurance products.
作者
王倩
邹全
WANG Qian;ZOU Quan
出处
《保险研究》
CSSCI
北大核心
2024年第7期51-63,共13页
Insurance Studies
关键词
心脑血管疾病
机器学习
住院费用预测
赔付预测
商业健康保险
cardio-cerebral vascular diseases
machine learning
prediction of hospitalization expenses
claim prediction
commercial health insurance
作者简介
王倩,北京工商大学经济学院风险管理与保险学系副教授,E-mail:qianwang@btbu.edu.cn;邹全,北京工商大学经济学院风险管理与保险学系硕士研究生。
