Visual Clustering of Supply Chain via Collaborative Annealing Power K-means++ Clustering

Efficient organization and management of supply chains are crucial for enhancing operational efficiency and responsiveness in today's complex market dynamics. In this paper, we address the large-scale supply chain clustering problem by using an efficient and robust clustering approach. The collaborative annealing power k. means++ algorithm is used to create a lower two-dimensional visual clustering map that effectively tackles supply chain clustering problems. The proposed method provides optimal solutions by decomposing large-scale supply chain problems into independent, smaller, and manageable segments. It simplifies decision-making by exploring similar clusters represented by neighboring branches resulting from the collaborative annealing power k-means++ algorithm. Unlike other approaches in the literature, the proposed method can also achieve well-ordered topological representations of various work order families, processed by clusters of supply units, along with information on hierarchical sub-cell formation identifiable from the visually navigable map based on the grouping efficacy measure. This paper details the experimental results on ten benchmark datasets, demonstrating the superior performance of the proposed approach compared to several baselines in terms of effectiveness indices.