Privacy Preserving Cross-Silo Federated Learning Frameworks for Collaborative Inventory Optimization in Decentralized Global Retail Supply Chains

Authors

  • Kumar Jaganathan Author
  • Dr. Janarthanam S Author

Keywords:

Federated Learning, Privacy Preservation, Inventory Optimization, Retail Supply Chains, Decentralized Learning, Cross-Silo Collaboration

Abstract

Federated learning enables multiple retail organizations to collaboratively optimize inventory management while preserving proprietary supply chain data. This paper introduces FedRetail-SecureOpt, a novel privacy preserving cross-silo federated learning framework designed specifically for decentralized global retail supply chains. Our approach implements differential privacy mechanisms with adaptive secure aggregation to protect sensitive inventory distribution patterns and demand forecasting models across heterogeneous retail networks. The proposed system architecture incorporates Byzantine-robust aggregation techniques to ensure model convergence despite potential adversarial participants. Evaluation on two real-world retail datasets demonstrates that FedRetail-SecureOpt achieves 94.7% accuracy in inventory demand prediction while maintaining differential privacy budgets of ε=1.2. We introduce two novel metrics: Cross-Silo Collaborative Efficiency (CSCE) and Privacy-Utility Trade-off Index (PUTI), which provide quantitative measures for federated learning performance in supply chain optimization. Results show 23% improvement in inventory turnover reduction and 31% decrease in stockout incidents compared to decentralized baseline approaches. The framework successfully demonstrates that privacy-preserving collaborative learning can significantly enhance retail supply chain resilience without compromising individual organizational data security.

Downloads

Download data is not yet available.

Downloads

Published

2026-09-22

How to Cite

Privacy Preserving Cross-Silo Federated Learning Frameworks for Collaborative Inventory Optimization in Decentralized Global Retail Supply Chains . (2026). Minimax Theory and Its Applications, 11(2s), 109-115. https://journalmta.com/index.php/jmta/article/view/402