Spare parts demand forecasting for mri devices based on AHP and K-means clustering

This study combines multi-criteria decision making and clustering methods to more accurately forecast spare part demands for magnetic resonance devices in the international market. The Analytical Hierarchy Process method is used to determine the relative weights of the criteria affecting spare part demands, and then the K-means clustering algorithm is applied to group countries based on similar demand characteristics. This approach particularly helps optimize spare part supply processes and improve demand forecasting accuracy in countries without manufacturing facilities. The study demonstrates that customized demand forecasts can be made for each cluster, thereby enhancing inventory management and supply chain processes. As a result, the developed model provides a significant contribution to improving the maintenance and repair processes of high-cost magnetic resonance devices in the healthcare sector and managing the spare part supply chain more effectively