Data mining techniques for intelligent educational management based on federated learning

As intelligent education grows quickly, the amount of data about managing education is also expanding quickly. Finding useful information quickly while keeping data private is the key to making education management smarter. This work presents an intelligent education management data mining technique grounded in federated learning (FL) and formulates a multi-level system architecture. The system can safely share and intelligently analyse dispersed education data thanks to improvements to the FL algorithm and the addition of the DP protection mechanism. Experimental validation using the PISA dataset indicates that the suggested approach markedly enhances accuracy, F1 Score, and AUC index of the model, while safeguarding data privacy and security, hence exhibiting better performance and robust generalisation capability. The study's findings furnish a theoretical foundation and technological assistance for advancing the evolution of informatisation and intelligence within the realm of educational management.