Leveraging Machine Learning to Identify Effective Teaching Strategies for Children With Autism Spectrum Disorder

Download PDF
Open Webpage

Savar Toteja, Pallavi Bajpai

Published: 2024, Last Modified: 05 Sept 2025COMPSAC 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Autism Spectrum Disorder (ASD) affects children globally, with rising prevalence underscoring the need for early and effective educational interventions. This study evaluates the use of machine learning (ML) algorithms to identify personalized teaching methods for children with ASD. We consolidated multiple datasets into a single, comprehensive dataset comprising 3043 samples with 17 features, which consists of behavioral and individual characteristics, further categorized into seven different teaching methods. We have used four ML algorithms: K-Nearest Neighbors (KNN), Multi-Layer Perceptron (MLP), Random Forest, and Decision Tree, to predict the most suitable educational strategies. Our results indicate that MLP and Decision Tree algorithms exhibited the highest accuracy in recommending tailored teaching methods. This study underscores the potential of machine learning in tailoring educational interventions for individuals with ASD, promising personalized support and enhanced learning experiences. Further real-world validation is necessary to optimize the effectiveness of this approach.