核密度法厘定我国粮食保险纯费率的实证研究  

An Empirical Study on the Net Premium Rates Calculation of Grain in China Based on Kernel Density Estimator

作　　者：梁来存[1] 

机构地区：[1]湘潭大学商学院,湖南湘潭411105

出　　处：《南京农业大学学报（社会科学版）》2009年第4期28-34,共7页Journal of Nanjing Agricultural University(Social Sciences Edition)

基　　金：国家社会科学基金项目(09BTJ016);国家统计局立项项目(2008LY035)

摘　　要：粮食保险属于政策性农业保险,准确的费率厘定将有助于政府制订出科学的保险支农政策。当样本较小、总体分布未知的情况下,传统的参数估计法厘定的费率不准确、不稳定。由于非参数核密度估计能够适用于任意分布,该文基于非参数核密度法探讨了粮食保险的纯费率厘定方法。并以高斯函数作为核函数,以S ilverm an的"经验法则"确定带宽,根据我国1979—2007年的实际数据,厘定了我国粮食单产保险的纯费率。Grain insurance belongs to policy-guided agricultural insurance. Accurate rate-making will be conducive to the policy-making of agricultural insurance. When samples are small and the total distribution is unknown, the calculated net premium rates are neither accurate nor stable using the traditional parametric methods. Because nonparametric kernel density estimator is suitable for any output distribution, the net premium rate cal- culation method is discussed based on nonparametric kernel density estimator. The net premium rates of grain crop output per unit area of all provinces are formulated with the use of Gauss function and Silvermans rule of thumb.

关 键 词：粮食保险 核密度估计 纯费率 费率厘定 

分 类 号：F326.11[经济管理—产业经济]