Multivariate Data Analysis for Customer Segmentation using Principal Component Analysis and K-Means Clustering

Abstract

This study aims to perform customer segmentation using Principal Component Analysis (PCA) and K-Means Clustering on a Mall Customers dataset consisting of 200 customers with attributes such as age, annual income, and spending score. The main problem addressed is the lack of understanding of diverse customer characteristics, leading to less effective marketing strategies. The proposed solution involves applying PCA to reduce data dimensionality and identify key features, followed by K-Means clustering to group customers into homogeneous clusters. The analysis results reveal three optimal clusters with distinct profiles: older customers with medium income and low spending scores, middle-aged customers with high income and moderate spending scores, and younger customers with low income but high spending scores. PCA visualization confirms clear segmentation, although some overlap exists near cluster boundaries. In conclusion, this segmentation provides valuable insights for companies to develop more targeted and effective marketing strategies based on the unique profiles of each customer cluster.