摘要
考虑每期索赔计数变量之间基于几何一阶整值自回归(NGINAR(1))相依结构的离散风险模型,利用下临界分支过程的大偏差,获得了一阶几何整值自回归过程的大数定律呈指数衰减,从而建立了有限破产概率的渐近表示.
We considered discrete-time risk model in which a dependent structure of new geometric first-order integer-valued autoregressive NGINAR（1） is introduced between the claim numbers for each period.With the help of large deviation of a sub-critical branching process,we derived law of large numbers is exponential decay for NGINAR（1） process,and then established the uniform asymptotic formula for the finite-time ruin probability.
出处
《数学的实践与认识》
北大核心
2016年第17期257-260,共4页
Mathematics in Practice and Theory
基金
齐齐哈尔市科学技术局软科学项目(RKX-201513
RKX-201403)
关键词
离散风险模型
几何整值自回归过程
大偏差
有限破产概率
discrete-time risk models
new geometric integer-valued autoregressive
large deviation
finite-time ruin probability
