The Analysis of Adult Autism Spectrum Disorders Screening Using Neural Network

Abstract

One of the increasing developmental disorders in Indonesia is Autism Spectrum Disorder (ASD), developmental disorder characterized by difficulties to conduct verbal and non-verbal communication and social interaction. This disorder cannot be tolerated and requires early treatment to reduce its development. However, ASD treatments required ineffective treatment costs and waiting times diagnosis were lengthly. One effective alternative diagnosis isto use the screening technology to determine the early symptoms of ASD disorders. The rapid development of the number of ASD cases around the world required researchers to determine a dataset with behavioral properties to update the screening process. Thus, the purpose of this study is to predict the success of screening performed on adults with Autism Spectrum Disorder (ASD) using the researchers’ results dataset, so that the dataset could be used as a benchmark for the success of the ASD screening process. The method used is machine learning neural network method with 100 training cycle, learning rate 0,01 and momentum 0,9 resulted in a classification accuracy of 96.00%