A stacking ensemble model for food demand forecasting: A preventative approach to food waste reduction

Abstract

Building effective demand forecasting is crucial for better planning and ensuring sustainability within food supply chain systems. The food industry has received the least attention for building demand forecasting approaches, with a noticeable lack of utilizing ensemble stacking models. Additionally, while some models have achieved accurate predictions, they do not consider freshness variables and are not assessed for their impact on waste reduction. This paper develops a demand forecasting framework that is considered as a preventative approach to reduce food waste by enabling food retailers to better manage inventory and balance supply with demand. The paper first develops an ensemble stacking model combining the random forest, support vector regression, eXtreme gradient boosting, long short-term memory models as base learners and Ridge regression as a meta-learner. The performance accuracy of the proposed model is assessed by benchmarking with singular models using various metrics. The experimental results reveal that the proposed stacking model outperforms random forest and eXtreme gradient boosting while consistently outperforming support vector regression and long short-term memory model, achieving a coefficient of determination score of 0.99, mean absolute error of 0.63, mean absolute percentage error of 1.8, and prediction accuracy of 98.2%. The model’s performance is further assessed on its impact on waste reduction by utilizing the predicted demand to replenish the inventory for the next day dynamically. The promising results indicate that relying on the predicted demand to replenish the inventory achieves a significant reduction in food waste.