A Kind Of Copula Based On The Zero-truncated Poisson Distribution And Its Application

As a tool for multivariate correlation analysis,Copula can capture the complex nonlinear relationships among variables,which are widely used in economics,insurance,engineering and other fields.Common types are Gaussian,t and Gumbel Copula.This paper promotes the work of Copula in common Copula.A new ZTP Copula is proposed based on the popular Copula,combined with max operator and ZTP(Zero-truncated Poisson)distribution.This paper firstly draw the conclusion that the new model is different from the common model under different parameters through graphing and K-L(Kullback-Leibler)distance.Then,the parameters of the model are estimated by semiparametric method.Simulation results show that as the sample size increases,the estimation results become more accurate.Secondly,the AIC(Akaike information criterion)is used to model the new model.The uncertainty analysis shows that the new model can still have a good estimation effect on the uncertain distribution.Finally,the use of worker compensation insurance data for real data analysis displays the application of the new model in real life.