Methods for Improving the Predictive Accuracy of Autism Spectrum Disorder Screening using Machine Learning Algorithms

Abstract

Autism spectrum disorder (ASD) is a developmental disorder that causes difficulties with communication and behavioral challenges especially in young children's. It is a psychological disorder characterized by restricted, repetitive behaviors and they are withdrawn from all forms of social interaction. ASD screening is normally conducted by a medical practitioner. The diagnosis of ASD is still made on the basis of behavioral observation. ASD screening tools have wide range questionnaire which aims to investigate whether child, adolescent and adult has symptoms of ASD or not. ASD screening involves gathering a large number of responses about the child or adult behavior either given by the parents or a caretaker. Autism screening process is considered to be lengthy and time consuming one as it relies on a simple domain expert, as well as a large number of questions that respondents have to answer. The correct diagnosis plays a crucial role in identification of autism. A possible way to improve the diagnosis accuracy and efficiency of the screening tools is to adopt intelligent methods based on machine learning algorithms. The aim of this paper is to automate the diagnosis process and enhance the predictive accuracy of the test, thereby speed up the screening process of autism spectrum disorder.