Food Demand Forecasting for Waste Reduction in Restaurants

Food wastage remains a persistent challenge in the restaurant industry, affecting both economic performance and the environment. This paper proposes the development of a machine learning-based system that accurately forecasts food demand, helping restaurants align production with actual consumption patterns. By analyzing historical sales, seasonal fluctuations, and external influences, the model aims to support more informed inventory planning. The anticipated outcome is a reduction in food waste, enhanced sustainability, and improved operational efficiency. The main objective of this investigation is to minimize food waste and improve operational efficiency through accurate demand prediction. The model utilizes advanced algorithms such as Random Forest and Long Short-Term Memory (LSTM) networks to capture non-linear consumption patterns and dynamic changes in customer behavior. The outcomes of this research indicate a significant reduction in overproduction and enhanced sustainability within restaurant management systems. From the study, it is concluded that the application of machine learning in food demand forecasting contributes to better resource utilization, cost savings, and alignment with global sustainable development goals related to waste reduction.