Computational Psychometrics: Machine Learning for Analyzing Learner Behavior in Educational Systems

Abstract

The integration of machine learning (ML) in educational technology has enabled deeper insights into learner behavior and cognitive patterns. This paper presents a computational psychometrics approach that leverages ML models to analyze educational data, predicting student performance, engagement, and learning preferences. We compare multiple ML algorithms, including decision trees, support vector machines, and deep learning models, to assess their effectiveness in identifying at-risk students and personalizing learning experiences. Empirical results on real-world educational datasets demonstrate the advantages of our approach in improving adaptive learning strategies. The study underscores the potential of ML-driven psychometrics in revolutionizing modern education.