Detection of Banks' Customers Loyalty Using Naïve Bayes and Support Vector Machine Classifiers: A Machine Learning Approach

Abstract

This research paper presents a machine learning approach for detecting and predicting customer loyalty in the banking sector. The study utilizes Naive Bayes and Support Vector Machine (SVM) classifiers to analyze customer data, including demographic information, transaction history, and customer feedback. The dataset is divided into training and testing sets for model development and evaluation. The Naive Bayes classifier leverages the assumption of feature independence, while the SVM classifier constructs optimal hyperplanes for class separation. Performance metrics such as accuracy, precision, recall, and F1-score are used to evaluate the models. Both classifiers demonstrate high accuracy in identifying loyal customers, indicating their potential for real-world application. The study also analyzes the influence of factors like age, income level, and transaction frequency on customer loyalty through feature importance analysis. The proposed machine learning approach offers valuable insights for banks to identify and target loyal customers, enabling effective customer relationship management and improved business performance. The research underscores the importance of feature engineering and model selection in developing ccurate customer loyalty prediction models.