Review of: "Bank Customer Churn Prediction Using SMOTE: A Comparative Analysis"

Résumé

Potential competing interests: No potential competing interests to declare.This work discusses the use of the Synthetic Minority Over Sampling Technique (SMOTE) and Genetic Algorithm (GA) in predicting bank customer churn.SMOTE helps address data imbalance by oversampling the minority class, improving model performance.GA is employed to select informative features from the dataset, enhancing the accuracy of churn prediction models.Various classification algorithms such as Random Forest, KNN, AdaBoost, and Artificial Neural Networks (ANN) are evaluated, with ANN showing consistent performance on both training and testing datasets.Results

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Voir l'article sur le site de la revue

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0