Machine Learning Model to Predict Autism in Children Using Facial Features
by Jasamarbir Singh Arora
Category: Computer Science
Abstract – Autism Spectrum Disorder (ASD) is a neurological condition resulting from differences in brain development. Detecting ASD in children is typically done using the Autism Diagnostic Interview-Revised (ADI-R), which can be expensive. To reduce the financial burden while providing a quick and easy diagnosis, I created an AI model that can identify ASD early in children based on their facial features. The model was trained with a dataset containing over 2000 images of Autistic and Non-Autistic children. 80% of this used for training and 20% for testing. Among the models tested, Logistic Regression demonstrated the best performance, an accuracy of 77%.