Article

Learning Analytics-Driven Student Performance Prediction Using Ensemble Machine Learning Models

---

Abstract

The increasing availability of data from digital learning environments has created new opportunities to analyse and predict learning performance using artificial intelligence. Learning analytics enables the extraction of meaningful behavioural patterns from temporal data to support data-driven educational decision-making. This study proposes learning analytics–driven framework for predicting short-term performance outcomes using ensemble machine learning models. The framework is evaluated on a dataset consisting of 1,334 temporal observations, integrating performance indicators, activity intensity measures, volatility metrics, and temporal lag features to capture both instantaneous and sequential learning behaviours. Five machine learning models—Decision Tree, Random Forest, Gradient Boosting, XGBoost, and a Voting Ensemble—are systematically compared using Accuracy, Precision, Recall, and F1-score as evaluation metrics. The experimental results show that the Decision Tree model achieves the highest F1-score of 0.546, while the Voting Ensemble model attains a balanced performance with an accuracy of 0.510 and an F1-score of 0.502, indicating stable classification behaviour across performance improvement and decline classes. Feature importance analysis reveals that short-term performance change indicators and recent activity intensity are the most influential factors in determining subsequent performance outcomes. Although predictive accuracy remains moderate due to the dynamic and volatile nature of temporal performance trajectories, the proposed framework demonstrates the potential of interpretable, learning analytics–driven ensemble models to support early performance tendency detection and adaptive learning interventions in dynamic learning environments.

Keywords: Learning Analytics, Ensemble Machine Learning, Student Performance Prediction, Temporal Behaviour Modelling, Artificial Intelligence in Learning, Ensemble Machine Learning, Student Performance Prediction, Temporal Behaviour Modelling, Artificial Intelligence in Learning

How to Cite:

Lestari, E. D., (2026) “Learning Analytics-Driven Student Performance Prediction Using Ensemble Machine Learning Models”, Artificial Intelligence in Learning 2(1). doi: https://doi.org//AIL.120