Abstract

Autism Spectrum Disorder is the general term given to a diverse group of neurodevelopment and brain disorders. The main symptoms of autism are relational interactive defects, verbal and non-verbal deficiencies of speech and repercussion. Electroencephalography is a method known for medical imaging as a detailed instrument for the analysis of signals produced by brain impulses and their behaviour. In this study, changes in brain EEG signals dependent on Recurrent Neural Network (RNN) was established to differentiate between normal and autistic children. The RNN achieves overall accuracy in classification of 92.7%.