Development of an Optimized Ensemble Least Squares Model for Identifying Potential Deposit Customers

Abstract

The banking sector faces significant challenges in effectively promoting its products and services. While direct marketing has proven to be a potent tool for customer acquisition, it often leads to customer dissatisfaction, thereby tarnishing the bank's reputation. Leveraging Business Intelligence (BI) technology offers a strategic advantage by enabling the classification and analysis of customer data, particularly for time deposit customers. This study presents the development and optimization of an Ensemble Least Squares (ELS) algorithm to enhance the classification of potential deposit customers. The proposed Ensemble Least Squares Support Vector Machine (ELS-SVM) algorithm demonstrated superior performance compared to traditional SVM and LS-SVM methods. Notably, the ELS-SVM achieved an average performance improvement of 10.04% over standard Support Vector Machine (SVM) techniques.

Keywords

• Business Intelligence
• Bank Marketing
• Classification
• Potential Deposits Customers
• Ensemble Least Square Support Vector Machine