Abstract: We considered the modeling problem of complex integer-valued time series data. Firstly, we  proposed a  class of first-order mixed integer-valued negative binomial autoregressive models, proved the strict stationary and ergodicity of the model, and discussed the probabilistic and statistical properties of the model such as transition probability, expectation, variance, etc. Secondly, we studied the  maximum likelihood estimation problem of the model, obtained the asymptotic normality of the estimator, and conducted empirical analysis on the basis of numerical simulations. The empirical analysis results show that the model performs well in fitting the drug offense count data.

Key words:  , integer-valued time series, mixed negative binomial autoregressive model, stationarity, maximum likelihood estimation