Computational Models for Academic Performance Estimation

TL;DR

This paper develops a data-driven, machine learning-based system for estimating student performance using a large, publicly available dataset, demonstrating superior accuracy over traditional methods.

Contribution

It introduces a large, publicly shared dataset and provides comprehensive analysis and comparison of deep learning and machine learning methods for performance estimation.

Findings

Deep learning approaches outperform traditional algorithms.

The dataset enables robust performance estimation across multiple courses.

The system works effectively with partially available student records.

Abstract

Evaluation of students' performance for the completion of courses has been a major problem for both students and faculties during the work-from-home period in this COVID pandemic situation. To this end, this paper presents an in-depth analysis of deep learning and machine learning approaches for the formulation of an automated students' performance estimation system that works on partially available students' academic records. Our main contributions are (a) a large dataset with fifteen courses (shared publicly for academic research) (b) statistical analysis and ablations on the estimation problem for this dataset (c) predictive analysis through deep learning approaches and comparison with other arts and machine learning algorithms. Unlike previous approaches that rely on feature engineering or logical function deduction, our approach is fully data-driven and thus highly generic with better performance across different prediction tasks.