Application and Evaluation of Multi-criteria Clustering Algorithms for Customer-Oriented Supply Chain Segmentation

Due to an ever increasing complexity, modern supply chains have to make use of Decision Support Systems to focus on the one hand on cost savings and on the other hand on a high customer orientation. This is especially difficult due to different service and costs expectations that have to be taken into account jointly to fulfill the customer expectations more precisely. While the need for a differentiated service fulfillment is generally acknowledged, little research exists to date addressing how an organization can identify different economical and logistical viable groups of a customer base for different logistical problem statements. Therefore, this paper seeks to enhance the knowledge in this area by the application and evaluation of a relative new data clustering algorithm based on Self Organizing Maps and comparing it to a standard data clustering algorithm based on the K-Means algorithm in the context of the research domain of supply chain segmentation.