This paper considers a new parameterization of the generalized binomial thinning operator that is to be incorporated in a simple ordered integer-valued autoregressive process(INAR(1))with the Poisson-Lindley innovatio...This paper considers a new parameterization of the generalized binomial thinning operator that is to be incorporated in a simple ordered integer-valued autoregressive process(INAR(1))with the Poisson-Lindley innovations.The statistical properties of the resulting INAR(1)process are explored along with the estimation procedures.Monte Carlo simulation experiments are executed to assess the consistency of the estimates under the new INAR(1)process.Finally,the importance of the proposed INAR(1)model is confirmed through the analysis of a real data set.展开更多
On the basis of a well-established binomial structure and the socalled Poisson-Lindley distribution,a new two-parameter discrete distribution is introduced.Its properties are studied from both the theoretical and prac...On the basis of a well-established binomial structure and the socalled Poisson-Lindley distribution,a new two-parameter discrete distribution is introduced.Its properties are studied from both the theoretical and practical sides.For the theory,we discuss the moments,survival and hazard rate functions,mode and quantile function.The statistical inference on the model parameters is investigated by the maximum likelihood,moments,proportions,least square,and weighted least square estimations.A simulation study is conducted to observe the performance of the bias and mean square error of the obtained estimates.Then,applications to two practical data sets are given.Finally,we construct a new flexible count data regression model called the binomial-Poisson Lindley regression model with two practical examples in the medical area.展开更多
文摘This paper considers a new parameterization of the generalized binomial thinning operator that is to be incorporated in a simple ordered integer-valued autoregressive process(INAR(1))with the Poisson-Lindley innovations.The statistical properties of the resulting INAR(1)process are explored along with the estimation procedures.Monte Carlo simulation experiments are executed to assess the consistency of the estimates under the new INAR(1)process.Finally,the importance of the proposed INAR(1)model is confirmed through the analysis of a real data set.
文摘On the basis of a well-established binomial structure and the socalled Poisson-Lindley distribution,a new two-parameter discrete distribution is introduced.Its properties are studied from both the theoretical and practical sides.For the theory,we discuss the moments,survival and hazard rate functions,mode and quantile function.The statistical inference on the model parameters is investigated by the maximum likelihood,moments,proportions,least square,and weighted least square estimations.A simulation study is conducted to observe the performance of the bias and mean square error of the obtained estimates.Then,applications to two practical data sets are given.Finally,we construct a new flexible count data regression model called the binomial-Poisson Lindley regression model with two practical examples in the medical area.
