A study of binomial AR(1) process with an alternative generalized binomial thinning operator

In order to describe the finite-range integer-valued time series data with dependent structure and excess zeros, we introduce a new binomial AR(1) process with an alternative generalized binomial thinning operator. Some probabilistic and statistical properties of this model are derived. Model parameters are estimated by conditional least squares method and conditional maximum likelihood method. The consistency and asymptotic normality of these estimators are studied. In addition, a real data analysis shows a better performance of the proposed model than other existing models.