考虑异常值的基于数学规划法的案均赔款法

发布时间：2018-06-03 15:18

  本文选题：案均赔款法 + 异常数据　； 参考：《财经理论与实践》2017年03期

【摘要】：在存在异常值的情况下,非寿险公司应用传统案均赔款法估计准备金常出现估计精度不高的问题。在分析传统案均赔款法中进展因子和结案率估计方法的缺陷之后,引入数学规划法对估计方法进行改进,建立进展因子和结案率的二次规划和目标规划模型,并引入权重因子弱化异常数据对进展因子和结案率估计结果的影响,运用一组流量三角形数据进行实证分析,结果表明:数学规划法可以有效估计进展因子和结案率,弱化异常数据对准备金估计结果的影响。
[Abstract]:In the case of outliers, the estimation accuracy of reserve is often not high when the traditional method of average compensation is used by non-life insurance companies. After analyzing the defects of the progress factor and the estimation method of the case closure rate in the traditional compensation average method, the mathematical programming method is introduced to improve the estimation method, and the quadratic programming and goal programming model of the progress factor and the completion rate are established. By introducing the weight factor to weaken the influence of abnormal data on the estimation results of progress factor and closure rate, an empirical analysis is carried out by using a set of traffic triangle data. The results show that the mathematical programming method can effectively estimate the progress factor and the completion rate. Weakens the abnormal data to the reserve estimate result influence.
【作者单位】： 山东科技大学数学与系统科学学院;
【基金】：国家自然科学基金项目(61502280)
【分类号】：F840.4
，

本文编号：1973245