Training& testing

This project investigates the use of machine learning algorithms for the task of supplier selection in supply chain analytics for Medizinn, a company producing medical and industrial equipment electronics. Supplier selection is important in order to achieve operational effectiveness and to deliver quality products, however, classical approaches towards supplier selection are often inadequate for the current supply chain lot of complexities and large amounts of data. This paper aims to address this issue of concern for decision support systems applications through the application of machine learning, specifically decision trees, random forests and support vector machines, to enhance and automate the process of supplier evaluation.

This research employs the historical supplier data collecting, features engineering and machine learning modeling to forecast a supplier based on such criteria as cost, quality, and reliability of delivery. The researcher trained and tested the models on supplier data to make predictions and recommendations. The findings of the study show that the application of machine learning does improve the importance and objectivity of the supplier selection process by minimizing human errors and time taken to make decisions. The accuracy of the predictive models was assessed by performance metrics including precision, recall and F1-score.

The discussion pertains to the arguments in favor of the use of machine learning such as more effective supplier assessment, exploitation of enormous amounts of data, and operational model development with time through new information. This project also mentions frustrations such as the quality of data and the difficulty in understanding advanced models of machine learning. Overall this project shows that using machine learning in supply chain analytics assists in making better supplier selection decisions, simplifying the procurement process, and improving risk management.

This study makes a case that machine learning can be embedded in the supply chain and offer relevant solutions that enhance the reliability, cost, and performance of the supply chain for example at Medizinn. The suggestions emphasize the need for suitable future research that integrates real time data with complex algorithms towards supplier selection.