Evaluating the Effectiveness and Impact of Machine Learning Models for Forecasting Meal Demand in University Restaurants
Resumo
Research Context: University restaurants face the challenge of balancing meal supply and demand, where efficient food management is essential to reduce waste and optimize costs, both crucial for service sustainability. Scientific and/or Practical Problem: Forecasting meal demand is complex due to consumption variability, the influence of holidays, and frequent menu changes. Proposed Solution and/or Analysis: This study applies machine learning models to forecast daily meal demand, using historical consumption data, holiday information, and menu attributes. The research also investigates the impact of these variables on forecast quality and accuracy. Related IS Theory: The study is based on Information Systems theory, particularly in the use of data and analytical intelligence to support decision-making. In this context, machine learning models act as information processors, capable of transforming real data into useful predictions for operational management. Research Method: The research followed the CRISP-DM framework: definition of objectives and context (Business Understanding); data collection and exploration (Data Understanding); cleaning and transformation (Data Preparation); training and calibration of models (Decision Tree, XGBoost, LSTM) (Modeling); performance evaluation with R² and comparison to the restaurant’s forecasts (Evaluation); and preparation for practical application (Deployment). Summary of Results: The XGBoost model achieved the best performance, reaching R²=0.92, outperforming the restaurant’s original forecast (R²=0.83). This represents a significant improvement in forecast accuracy, contributing to waste reduction and better inventory management. The inclusion of menu information was also shown to enhance prediction quality. Contributions and Impact to IS area: The study demonstrates how machine learning can be applied in institutional foodservice, providing operational, economic, and environmental benefits. It reinforces the role of Information Systems in enabling more accurate and sustainable decision-making.
Referências
Breiman, L. (2001). Random forests. Machine learning, 45(1):5–32.
Chapman, P. (2000). CRISP-DM 1.0: Step-by-step Data Mining Guide. SPSS.
Chen, T. and Guestrin, C. (2016). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, page 785–794, New York, NY, USA. Association for Computing Machinery.
Géron, A. (2022). Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow: Concepts, tools, and techniques to build intelligent systems. ”O’Reilly Media, Inc.”.
Hochreiter, S. and Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8):1735–1780.
Kanwal, N., Zhang, M., Zeb, M., Batool, U., khan, I., and Rui, L. (2024). From plate to palate: Sustainable solutions for upcycling food waste in restaurants and catering. Trends in Food Science Technology, 152:104687.
ONU (2024). Objetivos de desenvolvimento sustentável (ods). [link].
Pereira, H. A., Felix, L. B., Cordeiro, L. L., and Antunes, H. M. A. (2007). Predição do número de refeições utilizando redes neurais artificiais. In Anais do 8 Congresso Brasileiro de Redes Neurais, pages 1–5, Florianópolis, SC. SBRN.
Pereira, L., Ferreira, L., Nascimento, D., Vanderlei, I., Silva, D., Santos, V., Gomes, M., and Melo, I. (2025). Design, integration, and evaluation of a demand forecasting service in the context of primary healthcare. In Anais do XXI Simpósio Brasileiro de Sistemas de Informação, pages 506–514, Porto Alegre, RS, Brasil. SBC.
Rocha, J. C., Matos, F. D., and Frei, F. (2011). Utilização de redes neurais artificiais para a determinação do número de refeições diárias de um restaurante universitário. Revista de Nutrição, 24(5):735–742.
Rodrigues, M., Miguéis, V., Freitas, S., and Machado, T. (2024). Machine learning models for short-term demand forecasting in food catering services: A solution to reduce food waste. Journal of Cleaner Production, 435:140265.
Santos, Y. D. C. d. F., Saqui, D., and Dos Santos, P. C. (2021). Um método baseado em aprendizado de máquina para previsão da produção de refeições em restaurantes universitários: A method based on machine learning to predict meal production in university restaurants. In Proceedings of the XVII Brazilian Symposium on Information Systems, SBSI ’21, New York, NY, USA. Association for Computing Machinery.
United Nations Deputy Secretary-General (2024). School Meals Provide ’Powerful, Cost-effective Way’ to Drive Progress across Multiple Sustainable Development Goals, Deputy Secretary-General Tells Event. [link].
Yang, H., Jiao, W., Zouyi, L., Diao, H., and Xia, S. (2025). Artificial intelligence in the food industry: innovations and applications. Discover Artificial Intelligence, 5(1):60.
