The Application Of Zero-inflated Distribution In Short-term Individual Risk Model

In recent years, more and more research on the theory and application of zero-inflated data has been paid more and more attention by researchers. In agriculture, insurance actuarial science, economics, biomedicine, road safety, insurance and public health, and many other fields. There are a lot of technical data. Researchers often set up models to analyze such data by using binomial, negative binomial distributions, or Poisson distribution. But not the data can apply to ordinary model and real life count data often appear value is 0 with probability exceeds the scope of the traditional model, which is zero inflation. Obviously, if we continue to use the traditional model has not been able to accurately predict the scientific and research. Therefore, we must study the new model to be more consistent with the actual data of the law, so as to carry out deeper research and application.In this paper, we first give the reasons for the zero inflated Poisson distribution and some related problems, and the system is introduced in detail the form of zero inflated Poisson distribution. The parameters are estimated by using the method of maximum likelihood estimation, Bayesian method and moment estimation. The zero inflated Poisson distribution is introduced in detail.In the next section, taking into account the actual count data may exist value is 0, the probability of data is too high the phenomenon, we put zero inflated Poisson distribution is introduced into the insurance actuarial risk theory of short-term individual risk model to, by considering each individual insurance claim variables is zero inflated Poisson distribution of short-term individual risk model, conducts the research to the total claim distribution law, and the nature of short-term individual risk model are discussed.