Project 2: Building a Student Intervention SystemΒΆ

A pdf version is available here and the repository for the source of this document is here.

This project uses the UCI Student Performance data-set to model the factors that predict the likelihood that a student will pass his or her final exam. The project begins with an exploration of the data to understand the feature and target variables, followed by the selection of three machine-learning algorithms to potentially model the student-performance. Finally one of the models is chosen and the optimal parameters discovered using an exhaustive grid-search.