Comparative Evaluation of Machine Learning Models for Retail Sales Forecasting: A Multi-Algorithm Approach

Accurate sales forecasting is vital for retail operations, impacting inventory management and strategic planning. This study explores the application of advanced machine learning models- Linear Regression, Random Forest, XGBoost, Support Vector Regression, LSTM, and ARIMA to improve forecasting accuracy. A five-year dataset of retail transactions was preprocessed using differencing and lag features to enhance stationarity. Among all models tested, XGBoost demonstrated the highest predictive accuracy (R?: 0.989), outperforming traditional methods. The research provides a scalable framework for real-time forecasting, underlining machine learning's transformative role in the retail sector and offering practical implications for inventory optimization and customer demand prediction.