Prédiction du taux de désabonnement des clients d’un opérateur de télécommunications à l’aide de l’intelligence artificielle
African Journal of Development Informatics · 2026
DOI: 10.68173/ajdi.v1i1.3 ↗Kurz gefasst
Fachartikel in der Zeitschrift African Journal of Development Informatics, 2026. Verfasst von Massiré Kamara und 1 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This study examines customer churn prediction in the telecommunications sector using machine-learning techniques. The objective is to compare three models for identifying customers likely to cancel their services and thereby support retention strategies. The experiment uses a public Kaggle dataset containing 2,666 observations and 20 variables. Three models are evaluated: a decision tree, k-nearest neighbours, and a backpropagation neural network. In Python, the decision tree achieves an accuracy of 91.57% and an AUC of 84.59%, outperforming the neural network (87.64%; AUC 77.06%) and k-nearest neighbours (79.40%; AUC 63.33%). In Weka, the corresponding classification rates are 92.42%, 90.17%, and 89.79%. The decision tree therefore provides the best results among the three models within each evaluation protocol. Differences between Python and Weka should nevertheless be interpreted cautiously because the evaluation procedures are not identical. The study should be understood as a methodological proof of concept; validation with local operator data is required before operational use in Mali.
Zitieren
APA
Kamara, M., & Kante, M. (2026). Prédiction du taux de désabonnement des clients d’un opérateur de télécommunications à l’aide de l’intelligence artificielle. African Journal of Development Informatics. https://doi.org/10.68173/ajdi.v1i1.3
Vancouver
Kamara M, Kante M. Prédiction du taux de désabonnement des clients d’un opérateur de télécommunications à l’aide de l’intelligence artificielle. African Journal of Development Informatics. 2026. doi:10.68173/ajdi.v1i1.3
BibTeX
@article{Kamara2026,
author = {Kamara, Massiré and Kante, Mahamadou},
title = {Prédiction du taux de désabonnement des clients d’un opérateur de télécommunications à l’aide de l’intelligence artificielle},
journal = {African Journal of Development Informatics},
year = {2026},
doi = {10.68173/ajdi.v1i1.3}
}
EUVIMED