Transforming Inventory Management Through Machine Learning based Demand Forecasting for a Circular Economy

In this paper, we are investigating the challenges of demand forecasting for production in the food industry. Our research focuses on making machine learning models to improve demand forecasting in a circular economy, specifically for optimizing inventory management in a meal supply business. Accurately predicting demand can help reduce waste, enhance resource allocation, and promote sustainable practices. Our analysis has used various machine learning models in Python, including XGBoost, LightGBM, Random Forest, and Long short term memory(LSTM), and determined that XGBoost performed the best. We evaluated the models using RMSE, R-squared, and MSE metrics. By implementing these advanced forecasting techniques, businesses can optimize inventory levels, reduce waste, and make informed decisions to promote sustainability.