New York
LondonBeijing
A Machine Learning Approach for Detecting Autism Spectrum Disorder Using Classifier Techniques

A Machine Learning Approach for Detecting Autism Spectrum Disorder Using Classifier Techniques

Shilpi Bisht (Birla Institute of Applied Sciences, India) and Neeraj Bisht (Birla Institute of Applied Sciences, India)
Copyright: © 2022 |Pages: 21
DOI: 10.4018/978-1-6684-2443-8.ch001
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Autism spectrum disorder (ASD) is a neurological developmental disorder that results in infirmity in social behaviour and social communication. Autism is identifiable at any stage of life, but symptoms usually appear in the first two years. This chapter deals with ASD at three different levels: child, adolescent, and adult. For this purpose, the authors have used a dataset from the UCI repository submitted by Fadi Fayez Thabtah, which has 20 features. They proposed new supervised machine learning models to predict the possibility of autism disorder at the adult stage through child, adolescent, and adult datasets. The detailed comparative study of various methods developed is supported by performance measures and respective ROC curves.
Chapter Preview
Top

Introduction

Recently, Autism Spectrum Disorder (ASD) has pinched researchers' attention due to its fastest-growing nature and rapidly increasing cases almost all over the world. Autism spectrum disorder (ASD) is a lifelong neurodevelopmental disorder identified by some degree of disabilities in developing social-communicative skills, playing and imagination, and the presence of monotonous or constricted behaviours and interests (American Psychiatric Association, 2013; Ruzich et al., 2015). As per the WHO report (2019), one in 160 children has an autism spectrum disorder (ASD) (Elsabbagh et al., 2012).ASD usually begins in childhood and tends to persist into adolescence and adulthood. However, the conditions are most apparent during the first five years of life.

  • 1.

    Categories of Autism Spectrum disorder: Autism Spectrum disorder is categorized into five childhood-onset conditions. The three most common types are:

    • a.

      Autistic Disorder: It is the most common type of ASD whose symptoms are delay in speech, difficulty in social communication, impacting early hearing, learning disabilities, mental retardation and unusual behaviour.

    • b.

      Asperger Syndrome: This syndrome's symptoms are mild and do not involve learning disabilities, mental retardation, and speech delay. Patients generally show symptoms of difficulty in social communication and unusual behaviour.

    • c.

      Pervasive Developmental Disorder: Patients suffering from PDD involve only a few symptoms of either autistic disorder or Asperger syndrome but not all (Mythili & Shanavas, 2014).

  • 2.

    Diagnosis: Medical professionals like paediatricians or psychiatrists diagnose ASD by taking inputs from various disciplines. Standardized examination tools are at one's disposal, such as Screening Tool for ASD in kids and Younger Children “STAT: a 20 minutes monitoring for young children” and frequently explored “Autism Diagnostic Observation Schedule” (ADOS;16: 45 minutes monitoring done by an expert, available in different styles for persons of distinct degrees of language and ages, from one year to the adult stage) (Lord et al., 2018). Early intervention and diagnosis of ASD may help the subject get the necessary treatment and therapy in time, which certainly reduces the level of difficulties (hyperactivity, irritability, language and attention problems) faced by the patients. A detailed tabular representation of these screening and diagnostic instruments is given in table 1 below (MC, 2013):

Complete Chapter List

Search this Book:
Reset