Optimized E-Commerce Supply Chains: Federated Learning and Blockchain-Powered Decision Support

Abstract

As soon as supply chain management is done well in online marketplaces, it will lead to better logistics, lower costs, and satisfied customers. Customers are also happier. When Blockchain technology and Federated Learning (FL) are used together, they could make things safer. demand estimates that are accurate and the protection of personal information are both important. There is a lack of trust between stakeholders, data silos, and the possibility of cyberattacks, which are some of the difficulties that are associated with the supply chain management systems of today. By utilizing these models, individuals are able to make educated estimations regarding the amount of demand that will be there in the future. For the purpose of addressing these concerns, we suggest using a technique known as Blockchain-Based Decentralized Federated Learning (BC-DFL), which is a mechanism for forecasting demand. Using the immutable ledger that blockchain provides, this technology ensures the secure transportation of data. Having the ability to see demand in real time, providing members of the supply chain with increased trust, and requiring less assistance from third-party aggregators are just some of the many benefits that come with the solution that is proposed. Based on the findings, it is clear that the BC-DFL paradigm leads in an increase in demand. With five to ten nodes with fifty thousand to one hundred thousand data, the BC-DFL architecture may be able to achieve fifty rounds of proof-of-stake consensus and AES-256 encryption. The mean absolute error was 7.6, the root mean square error was 12.9, the R2 value was 0.91, and less than 5% of the data escape.