EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

The Poisson–QGamma Distribution: Properties, Estimation Methods, Regression Modeling, and Applications in Engineering Count Data

Fatma Zohra Seghier, Halim Zeghdoudi, Muhammad Ameeq, Sana Kanwal

Stats · 26.05.2026 · 9 (3) · 52

DOI: 10.3390/stats9030052 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Stats, online erschienen am 26.05.2026. Verfasst von Fatma Zohra Seghier und 3 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Modeling over-dispersed count data is a common challenge in applied statistics, especially in engineering applications where repeated events, system faults, and clustered observations often produce variability beyond that allowed by the classical Poisson model. In this paper, we introduce and study the Poisson–QGamma distribution, a new compound discrete model obtained by mixing the Poisson distribution with the QGamma distribution. The proposed distribution is analytically tractable and flexible enough to capture over-dispersion, skewness, and excess kurtosis, which are frequently observed in real count data. Several statistical properties of the distribution are derived, including the probability mass function, cumulative distribution function, survival and hazard rate functions, moments, dispersion index, skewness, kurtosis, entropy, and generating functions. Parameter estimation is considered using maximum likelihood, method of moments, least squares, and weighted least squares methods. The finite-sample behavior of these estimators is examined through Monte Carlo simulation. A regression model based on the Poisson–QGamma distribution is also developed for count responses with covariates. The proposed model is compared with classical and competing count models using simulation and real-data applications. Three engineering-related datasets, involving power grid failure counts, environmental sensor event counts, and packet loss counts in communication networks, are analyzed to illustrate the practical value of the model. The results show that the Poisson–QGamma model provides a better fit than several standard alternatives, including the Poisson, negative binomial, Poisson–Lindley, generalized Poisson, and COM–Poisson models, particularly in the presence of over-dispersion and heavy-tailed behavior. Overall, the proposed distribution offers a parsimonious and effective tool for modeling over-dispersed count data, while also contributing to the broader class of compound discrete distributions.

Zitieren

APA

Seghier, F. Z., Zeghdoudi, H., Ameeq, M., & Kanwal, S. (2026). The Poisson–QGamma Distribution: Properties, Estimation Methods, Regression Modeling, and Applications in Engineering Count Data. Stats, 9(3), 52. https://doi.org/10.3390/stats9030052

Vancouver

Seghier FZ, Zeghdoudi H, Ameeq M, Kanwal S. The Poisson–QGamma Distribution: Properties, Estimation Methods, Regression Modeling, and Applications in Engineering Count Data. Stats. 2026;9(3):52. doi:10.3390/stats9030052

BibTeX
@article{Seghier2026,
  author = {Seghier, Fatma Zohra and Zeghdoudi, Halim and Ameeq, Muhammad and Kanwal, Sana},
  title = {The Poisson–QGamma Distribution: Properties, Estimation Methods, Regression Modeling, and Applications in Engineering Count Data},
  journal = {Stats},
  year = {2026},
  volume = {9},
  number = {3},
  pages = {52},
  doi = {10.3390/stats9030052}
}
Kostenlos starten →