Advanced Performance Analysis in Sport Footwear Sales Prediction Using Machine Learning

Abstract—This research uses advanced machine learning techniques to analyse and predict sales trends in the sports footwear industry. It focuses on XGBoost, a state-of-the-art approach recognised for its predictive capabilities. The study aims to uncover critical performance drivers, enhance marketing strategies, and optimise sales efficiency. A thorough review of machine learning applications in sales forecasting within the footwear industry emphasises practical algorithms, innovative feature engineering, and external data integration to improve accuracy. Rigorous evaluation and comparison highlight XGBoost’s superior performance over Random Forests and Gradient Boosting, surpassing traditional forecasting methods in predictive accuracy. The research delivers actionable insights into sales performance across regions, product categories, and channels, empowering strategic inventory management and marketing decisions. By showcasing the potential of advanced analytics, this study significantly contributes to retail sales forecasting and provides a robust framework for future research and industry applications. Keywords - Sales forecasting, Sports footwear industry, Machine learning, XGBoost, Predictive analytics