Learning Style Prediction Using Artificial Neural Networks and Random Tree

Learners, comprising students, learn in distinct ways. Machine learning-based recognition of the learner’s style can inspire and advance academic performance. This study investigates the application of machine learning (ML) to recognize learners’ VARK learning styles in personalized and adaptive learning situations. The data for this investigation were gathered from 72 students in the Gifted Unit at Northern Border University, Saudi Arabia, utilizing a questionnaire that encompassed demographic and academic variables, and VARK responses. Two ML models (artificial neural networks and random tree) were built and assessed utilizing 10-fold cross-validation. Accuracy, mean absolute error, kappa statistics, ROC-AUC, and confusion-matrix analysis were used to measure model performance. The outcomes unveiled that both models classified learner-style groups with considerable performance, with the random tree (RT) model (Accuracy = 75.0%; kappa value = 0.5546) acting better than the artificial neural network (ANN) model (Accuracy = 73.61%; kappa value = 0.5356). The RT model also correctly categorized 54 of 72 cases, compared to 53 of 72 for the ANN model. These results uncover that ML tech