Predictive Analytics in Higher Education: A Systematic Literature Review of Machine Learning Models for Forecasting Academic Performance and Ethical Integration

This systematic literature review examines predictive analytics and machine learning (ML) applications in edu-cational settings from 2019 to 2023. Machine Learning (ML) algorithms are increasingly used to predict students' academic performance and are integral to early warning systems in educational settings. The study looks at several algorithms, from simple linear regression to more complex ensemble and neural network models. It also looks at how well they work in different types of schools and with different types of students. The analysis confirms that ML models provide educators with actionable insights to identify and support at-risk students. Along with evaluating predictive performance, the review discusses the ethical problems of using machine learning in education. These include worries about data privacy, the need for openness, and the possibility of bias in algorithmic decision-making. The paper concludes that while ML holds substantial promise in enhancing educational outcomes, its deployment requires prudent management to navigate the intersection of technological innovation and ethical practice, ensuring responsible use of data analytics in shaping the future of education.