| 简介： | Abstract: In this talk, I will review some recent progresses in modeling integer-valued time series data and mainly focus on two kinds of models: integer-valued autoregressive model and integer-valued GARCH model. Starting with motivations of these models, I discuss some fundamental issues, i.e. ergodicity and parameter estimation. A negative binomial integer-valued GARCH model is considered in details, real data analysis shows its superior performance compared with other models. |