Revolutionizing Retail: A Hybrid Machine Learning Approach for Precision Demand Forecasting and Strategic Decision-Making in Global Commerce

Abstract

A thorough comparison of several machine learning methods is provided in this paper, including gradient boosting, AdaBoost, Random Forest (RF), XGBoost, Artificial Neural Network (ANN), and a unique hybrid framework (RF-XGBoost-LR). The assessment investigates their efficacy in real-time sales data analysis using key performance metrics like Mean Absolute Error (MAE), Mean Squared Error (MSE), and R2 score. The study introduces the hybrid model RF-XGBoost-LR, leveraging both bagging and boosting methodologies to address the limitations of individual models. Notably, Random Forest and XGBoost are scrutinized for their strengths and weaknesses, with the hybrid model strategically combining their merits. Results demonstrate the superior performance of the proposed hybrid model in terms of accuracy and robustness, showcasing potential applications in supply chain studies and demand forecasting. The findings highlight the significance of industry-specific customization and emphasize the potential for improved decision-making, marketing strategies, inventory management, and customer satisfaction through precise demand forecasting.