This study aims to predict customer purchase behavior by classifying individuals into three categories: rare, occasional, and frequent buyers. Utilizing the "Customer Purchases Behaviour" dataset from Kaggle, we analyze key demographic and transactional features such as age, gender, income, education, region, loyalty status, purchase frequency, purchase amount, product category, promotion usage, and satisfaction score. These features provide crucial insights into customer behavior patterns. The primary objective is to build predictive models using Random Forest and XGBoost algorithms to classify the purchase behavior accurately. The dataset includes both categorical and numerical variables, allowing for a comprehensive evaluation of consumer patterns. Feature engineering, normalization, and model tuning are performed to optimize accuracy and performance. The classification outcome aids businesses in targeted marketing, loyalty strategies, and promotional planning. This predictive modeling approach supports data-driven decision-making in retail and e-commerce sectors by identifying high-value customers and tailoring engagement strategies effectively.

Keywords: Customer Purchase Behavior, Classification, Random Forest, XGBoost, Predictive Modeling, Consumer Analytics, Purchase Frequency, Retail Strategy, Data Mining, Machine Learning, Loyalty Status, Promotion Usage, Satisfaction Score.