For this competition our team decided to develop a tool which can predict a student’s grade in any course that he or she has not completed yet. Prior to starting we had to create a database with 14 students, 25 courses, and 150+ student-course relationships. This prediction will be made based on the results of other courses that this student has already finished. 
We load data from those completed courses and setup 4 metrics, each accounting for different academic aspect. Once the training set is established, we use scikit-learn multinomial logistic regression to train the model. Then the model can be used to predict grades on the test set. 

This tool can be used by students to know what grade should be expected in the class that they are enrolling it. Based on this prediction they can adjust, whether to work harder (if the prediction is pessimistic) or focus on other courses (if it is optimistic). The tool can also be used by professors to see how many students in their course are expected to need more help.

Share this project:

Updates