Intelligent educational data analysis with gaussian processes

As machine learning evolves, it is significant to apply machine learning techniques to the intelligent analysis on educational data and the establishment of more intelligent academic early warning system. In this paper, we use Gaussian process (GP)-based models to discover valuable inherent information in the educational data and make intelligent predictions. Specifically, the mixtures of GP regression model is adopted to select personalized key courses and the GP regression model is applied to predict the course scores. We conduct experiments on real-world data which are collected from two grades in a certain university. The experimental results show that our approaches can make reasonable analysis on educational data and provide prediction information about the unknown scores, thus helping to make more precise academic early warning.