Autism spectrum disorder diagnosis using optimal machine learning methods

Abstract

Autism spectrum disorder (ASD) is the disorder of communication and behavior that affects children and adults. It can be diagnosed at any stage of life. Most importantly, the first two years of life, regardless of ethnicity, race, or economic groups. There are different variations of ASD according to the severity and type of symptoms experienced by people. It is a lifelong disorder, but treatment and services can improve the symptoms. The literature focuses on one of the main methods used by physicians to diagnose ASD. Many types of research and medical reports have been reviewed; however, a few of them only give good medical results for the strong differentiation of ASD from healthy people. This paper focuses on using machine learning algorithms to predict an individual with specific ASD symptoms. The target is to predict an individual with specific ASD symptoms and finding the best machine learning model for diagnosis. Further, the paper aims to make the autism diagnosis faster to deliver the required treatment at an early stage of child development.