Towards a Sustainable Retail Food Chain: Artificial Intelligence Driven Dynamic Pricing and Promotions for Food Waste Reduction

Abstract

Food wastage represents a substantial environmental, economic, and social challenge, contributing to resource reduction. In response to these issues, this study investigates the potential of AI-driven dynamic pricing and promotions to decrease food waste within the supply chain of food retail. By leveraging machine learning models, namely, Gradient Boosting, SVR, Decision Tree, KNN, Random Forest, and Neural Networks, the research seeks to optimize pricing strategies for perishable goods based on factors such as shelf life, inventory levels, and demand fluctuations. The model’s performance was evaluated using MAE, R² scores and RMSE. Gradient Boosting emerged as the most effective model, achieving the lowest error rates (MAE: 0.113, RMSE: 0.536) and highest R² (0.828), indicating strong predictive power and accuracy in price adjustment. The results demonstrate that AI-driven dynamic pricing can accurately adjust prices in real-time, encouraging the sale of near-expiration items and thereby reducing food wastage. Future research may explore reinforcement learning approaches and expanded datasets to further refine pricing accuracy and expand the model’s applicability across different retail contexts.