<![CDATA[Inventory demand forecasting is vital for small and medium enterprises (SMEs) in the food manufacturing sector to maintain optimal stock levels, reduce waste, and improve operational efficiency. Traditional statistical methods often fail to capture complex demand patterns, necessitating the adoption of advanced machine learning (ML) approaches. This study conducted a comparative analysis of four ML models Long Short-Term Memory (LSTM), Facebook Prophet, XGBoost, and Gradient Boosting Regressor using a three-year dataset (January 2020–December 2022) from a Nigerian food manufacturing SME. The dataset included monthly demand records for thirteen product categories. Models were evaluated based on Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and R-squared (R²). LSTM consistently outperformed other models, achieving the lowest RMSE and MAE values and the highest R² scores, demonstrating superior capability in capturing non-linear and temporal demand patterns. Facebook Prophet and Gradient Boosting performed moderately, with Prophet offering higher interpretability. XGBoost showed the weakest predictive performance across all metrics. The findings indicate that LSTM is the most effective model for inventory demand forecasting in SMEs with dynamic demand profiles. Incorporating advanced ML techniques like LSTM can enhance forecasting accuracy and support strategic inventory management decisions in food manufacturing SMEs.]]>
Copyrights © 2025
