MACHINE LEARNING BASED CUSTOMER SEGMENTATION

Authors:

Ms. K. Kiranmai1, J. Pallavi2, N. Yamini3, V. Shri4, K. Jogendra5

^1,2,3,4,5MVGR College of Engineering, Vizianagaram, India

Abstract— Customer Retention is one of the major challenges faced by businesses in this modern world. The main goal and objective of this particular project is to develop a system known as "Customer Segmentation and Churn Prediction System" based on the concept of Machine Learning. The system would help businesses take decisions based on various Machine Learning algorithms such as Gaussian Mixture Model (GMM) and Spectral Clustering to segment the customers based on their behavior and other factors. Additionally, the system would also help businesses predict various outcomes based on other Machine Learning algorithms such as Extra Trees, AdaBoost, and CatBoost to predict the churn of customers based on the pattern and trend. The prediction would also be presented to the business in the form of Power BI diagrams, giving an idea about the trend and amount of customer churn. This would, in turn, help the business retain its customers, resulting in business growth.

Index Terms— Customer Retention, Customer Segmentation, Churn Prediction, Machine Learning, Gaussian Mixture Model (GMM), Spectral Clustering, Extra Trees, AdaBoost, CatBoost, Ensemble Learning, Predictive Analytics, Data Visualization, Power BI, Customer Behaviour Analysis, Business Intelligence, Data-Driven Decision Making.

KEYWORDS— Customer Retention, Customer Segmentation, Churn Prediction, Machine Learning, GMM & Spectral Clustering, Ensemble Models, Power BI, Predictive Analytics.