Early Autism Detection Through Vocal Pattern Analysis: A Machine Learning Approach Using Support Vector Machines

Posted: Oct 16, 2025

• Wei Zhang
• Maria Rodriguez
• Kenji Tanaka
• Fatima Al-Jaber

Abstract

This research investigates the efficacy of machine learning algorithms, specifically Support Vector Machines (SVMs), in detecting autism spectrum disorder (ASD) through vocal pattern analysis in young children. We collected and analyzed vocal samples from 250 children aged 2-5 years, including 125 diagnosed with ASD and 125 neurotypical controls. The study employed feature extraction techniques focusing on prosodic elements, spectral characteristics, and temporal patterns. Our SVM model achieved 87.2% accuracy in classification, with precision and recall rates of 85.8% and 88.4% respectively. The results demonstrate that vocal pattern analysis combined with machine learning provides a promising non-invasive screening tool for early ASD detection. This approach could significantly reduce the time between initial parental concerns and formal diagnosis, enabling earlier intervention and improved developmental outcomes.