Zero-Inflated Log-Linear Models: A Comparative Analysis of Prediction Accuracy with Other Zero-Inflated Count Models
Journal of Research in Applied Mathematics · 2025-10 · 11 (10) · 11-13
DOI: 10.35629/0743-11101113 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Research in Applied Mathematics, 2025-10. Verfasst von Kayode O.J Kayode O.J und 3 weiteren Autor:innen.
Abstract (Originalsprache)
Zero-inflated count data, characterized by an excess of zero counts and overdispersion, are common across fields such as health, ecology, and social sciences. Models like zero-inflated Poisson (ZIP) and zero-inflated negative binomial (ZINB) have been widely used to analyze such data. However, zero-inflated log-linear (ZILN) models, though less commonly employed, offer a flexible alternative for handling excess zeros and overdispersion. This paper compares the predictive accuracy of ZILN models with ZIP and ZINB models, using simulated and real-world datasets. Results indicate that ZILN models provide competitive or superior predictive accuracy, especially under high overdispersion and zero inflation. We discuss the implications of these findings for researchers and practitioners analyzing zero-inflated count data.
Zitieren
APA
O.J, K. O. K., S.A, O. S. O., S.K, A. S. A., & I.O, O. I. O. (2025). Zero-Inflated Log-Linear Models: A Comparative Analysis of Prediction Accuracy with Other Zero-Inflated Count Models. Journal of Research in Applied Mathematics, 11(10), 11-13. https://doi.org/10.35629/0743-11101113
Vancouver
O.J KOK, S.A OSO, S.K ASA, I.O OIO. Zero-Inflated Log-Linear Models: A Comparative Analysis of Prediction Accuracy with Other Zero-Inflated Count Models. Journal of Research in Applied Mathematics. 2025;11(10):11-13. doi:10.35629/0743-11101113
BibTeX
@article{OJ2025,
author = {O.J, Kayode O.J Kayode and S.A, Omotoso S.A Omotoso and S.K, Atanlogun S.K Atanlogun and I.O, Omosuyi I.O Omosuyi},
title = {Zero-Inflated Log-Linear Models: A Comparative Analysis of Prediction Accuracy with Other Zero-Inflated Count Models},
journal = {Journal of Research in Applied Mathematics},
year = {2025},
volume = {11},
number = {10},
pages = {11--13},
doi = {10.35629/0743-11101113}
}
EUVIMED