Integrated machine learning concept with XG booster and random forest framework for predicting purchase behaviour by online customers in e-commerce social networks

E-commerce - electronic commerce has been used by various companies in recent years to communicate and retain clients or customers. Different marketing strategies, such as digital promotion and online advertising, have been used and integrated with E-commerce systems. Normally, the advertiser will post the affiliated links to redirect the customers to the merchant. In the e-commerce business, lack of transparency in exhibiting the products and providing trust among the consumers in the e-commerce spaces are doubtful. In this research, a hybrid model with the integration of XGBoost and Random Forest classifier has been utilized to predict the current and previous behaviors of the consumers in order to identify the final action of whether the consumer will buy the product or not. This process mainly alleviates data overload issues and offers better-personalized services to targeted online consumers in various applications. The performance has been evaluated on certain metrics such as Accuracy, Precision, Sensitivity, Specificity, F1 measure, and AUC score. Eventually, the proposed hybrid model results were compared with the existing methods based on accuracy and procured 91.1%, which is effective compared to other existing algorithms.