Comparative Analysis of Adaptive Learning and Fast for Word Programs for ASD Students in Learning English, and Mathematics, and Predicting Future Academic Performance Using Machine Learning Algorithms

This study investigates the effi cacy of adaptive learning methods in teaching English and Mathematics to students diagnosed with autism spectrum disorder (ASD), compared to the Fast ForWord program. Utilizing a randomized controlled trial design, students aged 6-7 were assigned to either the adaptive learning group or the Fast ForWord group. Pre- and posttests in English and Mathematics, along with engagement and behavior checklists, were used to assess outcomes. We employed machine learning techniques, including Support Vector Machine (SVM), K Nearest Neighbor (KNN), Gaussian Process Regressor (GPR), and Logistic Regression (LR), to predict student scores and analyze the eff ectiveness of these educational interventions. Results indicate that the Gaussian Process Regressor (GPR) is the best for predicting students’ future grades, adaptive learning methods signifi cantly improved academic performance and engagement compared to the Fast ForWord program, suggesting a need for personalized educational strategies in ASD. These fi ndings have signifi cant implications for educators and policymakers seeking to enhance educational outcomes for students with ASD.