Performance Evaluation of a Multi-Level Approach to Predict Learning Styles In E-Learning System

Machine learning algorithms have been widely used for predicting learning styles in personalized e-learning systems.This paper evaluates effectiveness of a multi-level model using K-Means and Decision Trees algorithms to cluster learners into groups based on their characteristics and classify learners into the learning style dimensions of the Felder-Silverman learning styles model (FSLSM).Learner interaction data extracted from a Moodle Learning Management System (LMS) was preprocessed and used as input for K-Means clustering to group learners according to behavioural similarities.The resulting clusters were used to train and test a Decision Tree classifier that labelled each learner's preferred learning style based on the FSLSM.The model was evaluated using standard classification metrics, including accuracy, precision, recall, and F1-score.Evaluation results show that the proposed model achieved a 95% overall accuracy, with an emphasis on correctly identifying the learning style category across FSLSM dimensions, demonstrating the strong predictive performance of the proposed multi-level model in supporting automated learning style prediction.