摘要
In this paper, we propose several approaches to estimate the parameters of the periodic first-order integer-valued autoregressive process with period T (PINAR(1)(T)) in the presence of missing data. By using incomplete data, we propose two approaches that are based on the conditional expectation and conditional likelihood to estimate the parameters of interest. Then we study three kinds of imputation methods for the missing data. The performances of these approaches are compared via simulations.
- 出版日期2014
- 单位吉林大学