【作者】 李冉冉；

【导师】 卢仿先；

【作者基本信息】 湖南大学 ， 金融学， 2008， 硕士

【摘要】 为了适应我国社会主义市场经济的发展,建立健全社会保障体系是保障经济和社会稳定发展的必要条件。从1994年底,我国进行城镇职工基本医疗保险的综合改革试点开始,至2007年底,全国绝大部分地区组织实施了城镇职工基本医疗保险,参保人数达到1.8亿人,医疗保险制度在我国得到了很大的发展。然而,近年来的改革实践表明,我国的基本医疗保险制度仍然存在相当多的问题需要解决。随着计算机技术以及学科交叉的发展,数据挖掘技术被越来越多应用到保险的风险分析中。由于基本医疗保险覆盖面的扩大,以及建立年限的增加,基本医疗保险所积累的数据量积聚增加,为我们尝试将数据挖掘技术应用到基本医疗保险改革的研究中提供了必要的条件。本文应用了几种数据挖掘的算法分析了我国某市的数据挖掘样本,旨在研究数据挖掘技术在我国城镇基本医疗保险中的应用方式及其可行性。首先,使用决策树、神经网络等数据挖掘的基本算法,以某市的基本医疗保险数据为样本,分析了参保个人的年龄、性别、参保时间、月缴费基数以及参保个人所在企业规模等因素对基本医疗保险账户使用的影响及其影响的程度,以期找出隐含的有效信息;结论表明,对个人账户消费影响最大的因素是月缴费基数,其次是参保年限、年龄以及企业规模,性别几乎不会起到任何影响。其次,使用时序算法对各参保单位的账户使用情况进行回归以及预测,以此作为医疗保险监管部门对各参保单位的账户进行考察的依据。最后,在以上分析的基础上,结合我国城镇基本医疗保险的发展提出了完善基本医疗保险的建议。更多还原

【Abstract】 Under the socialist market economy, perfect social security is essential for a steady social and economic development. Chinese basic medical insurance for urban employees, which was firstly experimented at the end of 1994 had been implemented in most of the regions by the end of 2007 with the number of insureds up to 18 millions. However, the practice of basic medical insurance for urban employees reform in recent years shows that there are still a number of problems need to be solved. With the rapid development of computer technology and interdisciplines, data mining is more frequently applied to the risk analysis in insurance. Because of the extension of regions involved in basic medical insurance and the increase of years since the construction of the system, data size of basic medical insurance increases rapidly, this is a necessary condition for us to study how to apply the data mining technology in our basic medical insurance for urban employees.The present paper uses several algorithms of data mining to analysis the data sample of a certain city, aiming at studying the mode and the feasibility of application of data mining technology in urban basic medical insurance in China. Firstly, based on the analysis of the data sample of the basic medical insurance for urban employees of a giving city by using decision tree algorithm and neural network algorithm- two of the basic algorithms of data mining- this paper studies the effect of such factors as gender, age, years to be insured, the base of monthly payment of the insureds of urban basic medical insurance and the number of employees of the enterprise, to find out implicit effective information. The analysis comes to the conclusion that the base of monthly payment of the insureds is most important factor of the consumption of the personal account, then years of insured, age and number of employees of the enterprise are of less importance. More over, gender affects little. Then, it attemps to make regressions and predictions of the accounts of the insured enterprises by using time series algorithm, which will provide basis for the government supervisors’examination of the states of the accounts of the insured enterprises. Based on the above analysis, the paper gives relevant, in the light of the current condition of basic medical insurance for urban employees.更多还原