A comparative assessment of most widely used machine learning classifiers for analysing and classifying autism spectrum disorder in toddlers and adolescents

Abstract

Individuals with autism spectrum disorder (ASD) have social interaction and communication challenges due to a disruption in brain development that impacts how they perceive and interact with others. The symptoms of autism often present themselves within the first two years of a persons life but can be diagnosed at any time. The ASD theory suggests that the onset of symptoms occurs in early childhood and persists until young adulthood. Machine learning research has been applied to the study of autism. We use Machine Learning (ML) methods and technologies, including Naive Bayes (NB) classifiers, Logistic Regression (LR), Support Vector Machine (SVM), and Random Forest (RF) to analyse and predict ASD in toddlers (kids between 12 to 36 months old) and adolescents (children between 12 to 18 years old) using non-clinical ASD datasets. In this paper, we review the literature and conduct an in-depth investigation of supervised machine-learning algorithms to evaluate the effectiveness of four leading classifiers on a dataset used for ASD screening.