Reducing Food Waste in Culinary Institutions: A Predictive Model for Inventory Management Using Linear Regression

This study employs linear regression to enhance inventory management, integrating IoT-based data collection for real-time decision-making. By contrasting traditional heuristic models, the approach achieves a 10% improvement in forecast accuracy. Effective inventory management of perishable foods in culinary institutions is crucial for reducing waste, optimizing costs, and maintaining food quality. This research explores applying linear regression techniques for enhancing perishable food inventory management, aiming to predict demand, optimize stock levels, and reduce spoilage. By leveraging historical sales, inventory, and waste data, a linear regression model is developed to forecast perishable food consumption and provide actionable insights for inventory decision-making. In addition to linear regression, this study incorporates machine learning techniques such as decision trees and support vector machines (SVM) to compare predictive performance and identify the most effective approach for this specific use case. Unlike traditional inventory management methods, this study adopts a data-driven, predictive framework, moving beyond rule-based approaches by integrating machine learning models for more accurate and dynamic decision-making. The results demonstrate improved inventory accuracy and reduced waste, positioning the research as a significant advancement in utilizing machine learning for real-time, data-driven decisions in culinary institutions. This work builds upon existing studies by applying multiple Machine Learning models, showcasing their potential for further optimization in perishable food management systems.