Student Grade Prediction in Education Sector Using Machine Learning

Abstract

Great results are achieved by working more on weak zones rather than working on strong areas. Bringing the same picture in context of student exams, students will be able to achieve higher grades if they work on lagging areas. Hence, realization of student’s performing abilities is very important for both teacher and the student. One of the ways to do this is using student grade prediction. By using machine learning algorithms, we can predict how well the students are going to perform so that we can help the students whose grades are predicted low.

It has been observed that till now, almost all popular student grade prediction models have been made on foreign university’s datasets. This work proposes a student grade prediction model which uses Machine Learning and Deep Learning technologies for predicting Indian Student’s Marks. As India is a large country and it is impossible to sample students from the entire county, so the scope of the problem is narrowed down to the Computer Science Engineering students of popular colleges in Punjab (Ludhiana) area. A dataset of thousands of students has been formed by manual collection of information from the students via internet such as Google Forms and then has been processed for its data analysis, preceded by various machine and deep learning algorithms like Linear Regression, Ridge, Lasso, Elastic Nets, Random Forests, Gradient Boost, SVM, Deep Learning and Decision Trees.