The new Poisson weighted exponential minification INAR(1) process

Irshad, M. R., Aswathy, S., Maya, R.
OPSEARCH · Springer Nature · 2026-10-21 · DOI: 10.1007/s12597-026-01308-6

Аннотация

This paper presents the new Poisson-weighted exponential minification integer-valued autoregressive (NPWE-MINAR) process, motivated by the need to overcome the zero-absorbing property of conventional binomial and negative binomial thinning operators in discrete minification models and to provide greater flexibility for modeling overdispersed count time series. Various stochastic properties, particularly those related to conditional stochastic measures such as transition probabilities, the conditional probability generating function, conditional mean and variance, and the autocorrelation function, are analyzed. Two estimation methods are explored, and the accuracy of the resulting estimates is evaluated through simulation study. The practical usefulness of the NPWE-MINAR model was demonstrated through the analysis of a time series dataset, where it exhibited either superior or comparable performance to the conventional INAR model.

Читать у издателя →

Похожие статьи