Multilayer Perceptron-Based System for Behavioral Analysis and Educational Recommendation in Autism Spectrum Disorder

Abstract

Autism Spectrum Disorder (ASD) requires early identification and effective educational intervention. This study presents a multilayer perceptron neural network to identify behavioral patterns associated with ASD from questionnaire responses. The model achieved 100% accuracy, precision, recall, and F1-score on the test set (Ntest = 211). A rule-based interpretative layer translates model outputs into educational profiles and personalized pedagogical recommendations, enabling practical support for educators. The integration of machine learning with educational strategies demonstrates the potential of AI as a support tool for inclusive education and early intervention.