Evaluation of Machine Learning Models in Student Academic Performance Prediction
A.G.R. Sandeepa, Sanka Mohottala · 2025
Read Evaluation of Machine Learning Models in Student Academic Performance Prediction by A.G.R. Sandeepa, Sanka Mohottala (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-08047.txt.
Description
Forecasting which students will struggle is a classic task for machine learning, and Sandeepa and Mohottala compare standard models on school data mixing behavior, grades, and background details. A multi-layer perceptron, a small neural network, came out on top, reaching about eighty-six percent accuracy on held-out students. Averaged over ten validation rounds it scored near eighty percent, while memorizing the training set almost perfectly.
Careful feature selection did much of the work, and explainable methods were used to open the black box and check that the chosen features made sense. The authors argue that compact neural networks are data-efficient candidates for early-warning systems in schools.
Catalog record
| CREATOR | A.G.R. Sandeepa, Sanka Mohottala |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-08047