Claim Missing Document
Check
Articles

Found 1 Documents
Search

KOMPARASI REGRESI ZIP DAN REGRESI ZINB PADA DATA KEMATIAN BAYI DI JAWA BARAT Fitri, Dina Jumiatul; Siregar, Anggrainy Togi Marito; Haay, Happy Alyzhya; Samosir, Rasi Kasim
Ensiklopedia of Journal Vol 8, No 1 (2025): Vol. 8 No. 1 Edisi 2 Oktober 2025
Publisher : Lembaga Penelitian dan Penerbitan Hasil Penelitian Ensiklopedia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33559/eoj.v8i1.3439

Abstract

Poisson regression analysis is a nonlinear regression that is usually used for discrete data and assumes equidispersion. In practice, there is often a violation of assumptions equidispersion, one of the violations is overdispersion (the variance is greater than the average value). One of the causes of overdispersion is the excessive number of zero values (Excess Zero) on the response variable. Excess zeros can be seen in the proportion of response variables which is zero greater than any other discrete data. There are many methods to overcome overdispersion: Zero Inflated Poisson (ZIP) regression and Zero Inflated regression Negative Binomial (ZINB). The purpose of this study is to determine the regression model which is better used on data that experience overdispersion. Data used to analyze ZIP and ZINB regression is the data of infant death in West Java Province in 2021. Based on the study’s result, it is known that the Akaike Information Criterion (AIC) value in the ZINB regression is smaller than the ZINB regression AIC value. So the ZINB regression is better used as well as the factor of infant death.Keywords: Overdispersion; Poisson Regression; ZIP; ZINB; Infant Death