Enhancing Bank Marketing Strategies: The Impact of Feature Reduction Techniques on Machine Learning Model Performance

Abstract

This research investigates the application of machine learning models in the banking sector, specifically focusing on the classification of bank marketing datasets. We explore the use of feature reduction techniques, including Principal Component Analysis (PCA) and SelectKBest, to enhance the performance of various models such as Multi-Layer Neural Networks (MLNN), K-Nearest Neighbors (KNN), Random Forest(RF), and Linear Discriminant Analysis (LDA). Our study reveals the pivotal role of these reduction techniques in addressing the challenges posed by high-dimensional data and imbalanced distributions in customer behavior prediction. Through comparative analysis, we demonstrate how PCA and SelectKBest, through F-Value and Chi-Squared methods, influence model accuracy and efficiency, providing insights into the effectiveness of these models in strategic bank marketing efforts.