Predicting Retail Supply Chain Disruptions Using Graph-Based Machine Learning Models

Vemula, Shashidar (2025) Predicting Retail Supply Chain Disruptions Using Graph-Based Machine Learning Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (788kB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

The research problem reported in this study is the prediction of disruptions in retail supply chains using graph-based machine learning in an effort to overcome the weaknesses of the conventional models, since they assume that supply-chain entities are autonomous units. The study uses transactional data to create a customer-product bipartite graph and compares three Graph Neural Network variants, namely GCN, GAT, and GraphSAGE to the Random Forest and SVM models. Extensive experiments prove that although GNNs are able to learn relational structure, their quality is limited due to a lack of graph-relevant information and a lack of class representation. Random Forest is better predictive accuracy and interpretable, which the SHAP analysis had validated, suggesting that node-feature disruption is the primary cause of this dataset. The results point out both the potential and existing constraints of graph-based methods, and provide a basis upon which more network-aware disruption forecasting can be realized in the future.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Sahni, Anu UNSPECIFIED
Uncontrolled Keywords:	Supply Chain Disruption; Graph Neural Networks; Graph Convolutional Network; Graph Attention Network; Explainable AI; Retail Analytics
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Retail Industry H Social Sciences > HD Industries. Land use. Labor > Business Logistics > Supply Chain Management
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	09 Sep 2026 10:42
Last Modified:	09 Sep 2026 10:42
URI:	https://norma.ncirl.ie/id/eprint/9924

Actions (login required)

View Item