Global supply chains have become increasingly complex, and ensuring reliable delivery has become a major challenge for logistics and operations managers. As the supply chain networks increase their scope and cover more regions and various stakeholders, the capacity to forecast delivery risk before it occurs has become essential for maintaining the supply chain’s resilience and the service’s consistency.Conventional risk assessment techniques tend to be based on reactive monitoring and limited analytical capabilities, which may not be able to identify the complex interaction between operational, logistical, and external risk factors. In recent years, machine learning has emerged as a promising approach for extracting predictive insights from large and heterogeneous supply chain datasets.This study investigates the effectiveness of several machine learning algorithms in predicting delivery risk within global supply chains.This study utilizes structured shipment data that contains operational and transportation details, as well as outside risks, to create and subsequently evaluate using supervised models including logistic regression, naïve Bayes, support vector machines (SVMs), decision trees, random forests, and gradient boosting. A systematic experimental framework is implemented to ensure consistent model training and evaluation, with performance assessed using commonly adopted classification metrics such as accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (ROC–AUC). The findings reveal meaningful differences in model performance, with ensemble-based approaches demonstrating stronger predictive capability in capturing complex interactions among supply chain variables. The analysis also identifies several operational and contextual factors that play a critical role in shaping delivery risk.The research demonstrates how machine learning methods work in predictive risk assessment through its testing of various algorithms within a single testing framework. The research findings enable organizations to improve their supply chain risk management through data-driven approaches while establishing delivery reliability and building resilient global supply chains.
Acharya et al. (Sat,) studied this question.