Enhancing Supplier Selection and Order Allocation Processes with Machine Learning Clustering and Optimization Techniques

Optimizing supply chain management relics on effective Supplier Selection and Order Allocation (SS&OA). This study introduces a novel two-phase framework that exemplifies a sophisticated approach to SS&OA, demonstrating the efficacy of integrating modern data-driven methodologies to enhance decision accuracy. In the initial phase, K-means Gaussian Mixture Model, and Balance Iterative Reducing and Clustering techniques are utilized to identify and group suitable suppliers based on managerial preferences and specific requirements. The accuracy of the clustering models is assessed through the Silhouette Score technique. The second phase introduces a state-of-the-art multi-objective optimization model for SS&OA, strategically employing suppliers shortlisted from the most effective Machine Learning (ML) clustering method in Phase 1. The mathematical model has been developed encompassing considerations for multi-source, multi-period, and multi-product scenarios. Validated with real historical contract data from Canada, this framework highlights ML clustering techniques' substantial impacts on decision accuracy, offering valuable insights for data-driven SS&OA decisions.