Abstract
This study provides an up-to-date evaluation of predictive models used in forecasting student performance and learning outcomes in the educational sector. It examines modern machine learning and deep learning models for predicting college dropouts, utilising cost-effective large-scale administrative data as an early warning system. Additionally, the study investigates strategies for ensuring data privacy through federated and split federated learning techniques in forecasting dropout rates across a diverse student population. Furthermore, it addresses the importance of safeguarding data privacy, including student performance, user profiles and personalised learning paths, within the educational sector. The chapter encompasses big data analysis in education, exploring patterns, highlighting research themes, identifying limitations and proposing future directions. It also delves into the application of splitfed learning techniques to the prominent learning analytics challenge of predicting student dropouts, thereby contributing to both predictive accuracy and data privacy in educational settings.
Original language | English |
---|---|
Title of host publication | Split Federated Learning for Secure IoT Applications |
Subtitle of host publication | Concepts, frameworks, applications and case studies |
Publisher | Institution of Engineering and Technology |
Pages | 151-167 |
Number of pages | 17 |
ISBN (Electronic) | 9781839539466 |
ISBN (Print) | 9781839539459 |
DOIs | |
Publication status | Published - 01-01-2024 |
All Science Journal Classification (ASJC) codes
- General Computer Science