<![CDATA[Customer purchase behavior is a crucial factor in the development of effective marketing strategies. By leveraging predictive analytics, businesses can personalize recommendations, optimize marketing campaigns and improve user experience, ultimately contributing to increased conversion rates and customer retention. This research compares the performance of AdaBoost and CatBoost algorithms in predicting customer purchase behavior. The dataset used includes demographic attributes and customer behavior history, allowing for comprehensive analysis. The results showed that CatBoost performed better overall with an accuracy of 94%, while AdaBoost showed higher recall and F1-score values in the positive class. This study concludes that both algorithms have reliability in predicting customer behavior, where CatBoost is superior in handling categorical features, while AdaBoost offers good adaptability on simpler datasets. As a next step, future research can explore the implementation of these models in real-time scenarios.]]>
Copyrights © 2026
