A machine learning-based framework for optimizing emission factors in supply chain management within management information systems

As a component of a Management Information Systems (MIS), the current study proposes a machine learning-based solution for supply chain management emission factor optimization. Using advanced regression and clustering algorithms, the framework would be able to anticipate and categorize greenhouse gas (GHG) emissions, allowing firms to decrease their environmental impact and enhance operational performance. With an R² score of 0.914307 and a low Mean Absolute Error (MAE) of 0.145762, Random Forest Regression, in particular, beat Linear Regression in terms of accuracy when used to predict emission components. Industries were categorized according to their emission patterns using the K-Means and DBSCAN clustering algorithms; DBSCAN performed better (Silhouette Score = 0.857191). By incorporating these models into the MIS, supply chain emissions may be optimized via data-driven, real-time decision-making. Although these findings are impressive, incorporating feature engineering and other clustering approaches could further enhance the framework. The approach holds a great deal of promise for future advancements in the fields of multi-objective optimization and real-time data analysis because it sets the groundwork for a long-term, data-based solution to the global emissions challenge.