Optimizing Segmentation and Purchase Forecasting in Credit Card Transactions: A PSO-enhanced k-means and ANN Approach

Abstract

In the rapidly evolving landscape of data-driven marketing, machine learning has emerged as a pivotal tool for analyzing complex consumer behaviors and enhancing strategic decision-making. This paper introduces a novel approach to optimize customer segmentation and purchase forecasting in credit card transactions through the synergistic integration of Particle Swarm Optimization (PSO)-enhanced k-means clustering and Artificial Neural Networks (ANN). The proposed methodology refines customer segmentation by leveraging PSO, resulting in more defined clusters. In the predictive modeling phase, an ANN outperforms conventional methods, providing superior accuracy in purchase forecasting. The study demonstrates the effectiveness of advanced algorithms in enhancing insights from credit card transaction data, offering valuable implications for improved decision-making in the financial domain.