Hybrid Split-Federated Learning Framework for AI Driven Protection of Student Fee Payment Systems in Higher Education Institutions

DOI

https://www.doi.org/10.59256/indjcst.20260501055

Published In

Volume 5, Issue 1 (2026)

Published Online

January-April 2026

Pages

492-495

Article Type

Original Article

Check for updates

n the technological world, the usage of digital payment systems is increased in colleges and universities. This has increased the risk of financial data breaches, credential theft, and improper access to student databases in educational organizations. Traditional security monitoring methods will put sensitive financial records at greater risk due to storage and processing of transactional logs in one central place. This paper presents a Hybrid Split-Federated Learning (HSFL) framework for AI-driven mobile security testing, aimed at protecting student payment systems while maintaining privacy. The framework merges split learning and federated learning to allow for threat detection without sending raw financial data between devices or institutions. Login behaviors, API requests and transaction anomalies are examined in the initial model layers function on local devices. After this process, the intermediate activations are securely sent to institutional edge servers for joint model training and aggregation. This framework also includes secure aggregation methods that follow guidelines from OWASP to reduce data leak risks. The HSFL approach provides a scalable, privacy-focused, and strong solution for securing digital payment systems in today’s smart campuses.