Customer Segmentation Using Machine Learning Methods

Xin Lu

Abstract

This project aims to segment our customers by applying five different machine learning algorithms: k-means, DBSCAN, hierarchical clustering, Gaussian mixture, and spectral clustering. Through the collection of customer data and in-depth data analysis, we have gained critical insights into customer behaviors and preferences. By comparing these diverse segmentation methods, we have determined which approach is most suitable for our dataset and business needs. Ultimately, we provide market strategy recommendations based on the segmentation results and validate their effectiveness through customer behavior prediction. This project will facilitate a better understanding of our customers, enabling the provision of personalized products and services, ultimately leading to increased customer satisfaction and improved business outcomes.