Semi-Parametric Modeling of Churn Confounding Competing Risks Using Time-Dependent Covariates Among Mobile Phone Subscribers in Kenya

Ndilo B. Fwaru (), Leonard K. Alii () and Jerita J. Mwamb ()
Additional contact information
Ndilo B. Fwaru: Pwani University, School of Pure and Applied science, Dapartment of Mathematics and Computer Science, Kilifi, Kenya
Leonard K. Alii: Pwani University, School of Pure and Applied science, Dapartment of Mathematics and Computer Science, Kilifi, Kenya
Jerita J. Mwamb: Pwani University, School of Pure and Applied science, Dapartment of Mathematics and Computer Science, Kilifi, Kenya

EKOIST Journal of Econometrics and Statistics, 2023, vol. 0, issue 38, 75-86

Abstract: Mobile phone service providers are currently experiencing high churn rates. As a result, service providers are trying to develop ways to predict churn rates and uncover why subscribers’ churn occurs. However, the task of predicting churn in the mobile phone industry is complicated due to the large, sparse, and unbalanced nature of the data especially when competing risks are confounded by time-dependent covariates. This paper aimsto develop a semi-parametric model (the adjusted Cox model) by adjusting the extended Cox proportional hazards model to model competing risks confounded by time-dependent covariates and uses data from three mobile phone service providers in Mombasa and Kilifi Counties in Kenya to analyze and evaluate the validity and performance of the model. The paper establishes that the adjusted Cox model is a better model for predicting subscriber’s survival outcomes as well as for detecting the most influential covariates when competing risks are confounded with time-dependent covariates.

Keywords: Semi-parametric; Competing Risks; Time-dependent Covariates; Churn (search for similar items in EconPapers)
Date: 2023
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://cdn.istanbul.edu.tr/file/JTA6CLJ8T5/45BA768193C7423990D31AC8A0F31518 (application/pdf)
https://iupress.istanbul.edu.tr/tr/journal/ekoist/ ... subscribers-in-kenya (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ist:ekoist:v:0:y:2023:i:38:p:75-86

DOI: 10.26650/ekoist.2023.38.1159543

Access Statistics for this article

EKOIST Journal of Econometrics and Statistics is currently edited by Aycan HEPSAĞ

More articles in EKOIST Journal of Econometrics and Statistics from Istanbul University, Faculty of Economics Contact information at EDIRC.
Bibliographic data for series maintained by Istanbul University Press Operational Team (Ertuğrul YAŞAR) ().