An application of the generalized linear model for the geometric distribution

Farzana Jahan; Borhan Siddika; M. Ataharul Islam

doi:10.18642/jsata_7100121695

Back

An application of the generalized linear model for the geometric distribution

Journal article

Open access

Peer reviewed

An application of the generalized linear model for the geometric distribution

Farzana Jahan, Borhan Siddika and M. Ataharul Islam

Journal of Statistics: Advances in Theory and Applications, Vol.16(1), pp.45-65

2016

DOI: https://doi.org/10.18642/jsata_7100121695

Files and links (1)

pdf

Published232.47 kBDownload View

Open Access

Abstract

Statistical modelling

Statistics

Application of the generalized linear models (GLMs) in real life problems are well established and has extensive use. However, the GLM for the geometric distribution is not explored yet. The present study consists of the derivation of the GLM for the geometric distribution, estimation of parameters, and test procedures. An application is made to Bangladesh Demographic and Health Survey 2011 data to find the significant factors associated with the first occurrence of infant death in terms of birth order. Two different generalized linear models are fitted, one using the natural link function and the other one using the log link function. At the end, the results of both models are compared. It is found that the model fitted using log link function has lower Akaike’s information criteria (AIC) and deviance than the model fitted using the natural link function, that means the GLM for the geometric distribution using log link function provides better result.

Details

Title: An application of the generalized linear model for the geometric distribution
Authors/Creators: Farzana Jahan (Author) - Murdoch University, College of Science, Technology, Engineering and Mathematics
Borhan Siddika (Author) - University of Dhaka
M. Ataharul Islam (Author) - University of Dhaka
Publication Details: Journal of Statistics: Advances in Theory and Applications, Vol.16(1), pp.45-65
Publisher: Scientific Advances Publishers
Grants: Analysis of Repeated Measures Data under Higehr Education Quality enhancement project, Window 2 (Research), implemented by University Grants Commission, Bangladesh, CP-3293, World Bank
Identifiers: 991005575770107891
Murdoch Affiliation: College of Science, Technology, Engineering and Mathematics
Language: English
Resource Type: Journal article

Metrics

45 File views/ downloads

89 Record Views