Deep Classifier Model for Autism Spectrum Disorder Prediction

Abstract

Autism Spectrum Disorder (ASD) is a neurological and developmental disorder that affects human communication and behavior. ASD is associated with significant healthcare costs for diagnosis as well as for treatment. Disease diagnosis using deep learning model has become a wide research area. This paper proposes a deep classifier model for ASD prediction. The evaluation of the proposed model is performed over three datasets involving child, adolescent, and adult provided by ASDTest database. The obtained results showed that deep classifier model provides better results than other common machine learning classification techniques, with an accuracy of 99.50%, 99.23% and 99.42% for respectively adult, adolescent, and child datasets. Practical experiments conducted over these datasets report encouraging performances which are competitive to other existing ASD prediction models.