Herawati, Netti and Nisa, Khoirin A Robust Procedure for GEE Model. Far East Journal of Mathematical Sciences (FJMS). ISSN 0972-0871 (Submitted)

[img] Text

Download (139kB)


Abstract In longitudinal studies, multiple measurements are taken on the same subject at different points in time. Thus, observations for the same subject are correlated. This paper proposes a robust procedure for estimating parameters of regression model when generalized estimating equation (GEE) applied to longitudinal data that contains outliers. The procedure is a combination of the iteratively reweighted least square (IRLS) and least trimmed square (LTS) methods and is called iteratively reweighted least trimmed square (IRLTS). We conducted a simulation study for gamma model and Poisson model using the proposed method, the result shows that our approach can provide a better result than the classical GEE. 2010 Mathematics Subject Classification: 62J12. Keywords: longitudinal data, outlier, regression model.

Item Type: Article
Subjects: H Social Sciences > HA Statistics
Divisions: Fakultas Matematika dan Ilmu Pengetahuan Alam (FMIPA) > Prodi Matematika
Depositing User: Dr NETTI HERAWATI
Date Deposited: 07 Jun 2017 09:22
Last Modified: 07 Jun 2017 09:22
URI: http://repository.lppm.unila.ac.id/id/eprint/2972

Actions (login required)

View Item View Item