Education Data Analysis using Ensemble Models

In this evolving era of machine learning, educational data mining has immense potential in transforming the pedagogy methods. In higher educational institutes, improvement of students' performance is an important aspect for NIRF parameters such as Teaching and Learning. It is a crucial task to identify significant attributes present in the education dataset that contribute most to academic performance prediction. Such work helps educational stakeholders in developing better methods which will improve the effectiveness of Teaching and Learning. In this paper, various classifier models along with ensemble learning models have been used to compare performance of the students. The novelty of this work is to observe the effects of feature engineering on the classification accuracy of the model when applied to the educational dataset. The results obtained show that the Ensemble algorithms perform better than the base algorithms when making student grade predictions.