Predicting Students Performance in Educational Data Mining

TLDR

Educational Data Mining uses machine learning to predict student academic performance, but measuring performance is difficult due to diverse, nonlinear interrelationships among factors, making traditional data mining techniques inadequate. The study develops a deep‑learning classification model that automatically learns multiple representation levels to predict student performance. The model is built by layer‑wise pre‑training with a sparse auto‑encoder on unlabeled data, followed by supervised fine‑tuning on a large real‑world student dataset. Experimental results demonstrate the method’s effectiveness and its potential use in academic pre‑warning systems.

Abstract

Predicting student academic performance has been an important research topic in Educational Data Mining (EDM) which uses machine learning and data mining techniques to explore data from educational settings. However measuring academic performance of students is challenging since students academic performance hinges on diverse factors. The interrelationship between variables and factors for predicting performance participate in complicated nonlinear ways. Traditional data mining and machine learning techniques may not be applied directly to these types of data and problems. In this study we develop a classification model to predict student performance using Deep Learning which automatically learns multiple levels of representation. We pre-train hidden layers of features layerwisely using an unsupervised learning algorithm sparse auto-encoder from unlabeled data, and then use supervised training for finetuning the parameters. We train model on a relatively large real world students dataset, and the experimental results show the effectiveness of the proposed method which can be applied into academic pre-warning mechanism.