Dublin Core
Title
Machine Learning in Autism Spectrum Disorder Diagnosis
Abstract
This paper represents an overview of Machine Learning techniques used in Autism Spectrum
Disorder - ASD diagnosis. ASD is detected based on behavioral screening which is time consuming and
can only be taken by a medical professional. The idea is to find a smaller number of features that are still
able to equally well provide satisfying results and not lose the accuracy, sensitivity nor specificity. Some
of the algorithms mostly used in recent studies were Artificial Neural Network - ANN and Alternating
Decision Trees - ADTrees. The researches usually use WEKA software package for applying the algorithm
and obtaining results.
Disorder - ASD diagnosis. ASD is detected based on behavioral screening which is time consuming and
can only be taken by a medical professional. The idea is to find a smaller number of features that are still
able to equally well provide satisfying results and not lose the accuracy, sensitivity nor specificity. Some
of the algorithms mostly used in recent studies were Artificial Neural Network - ANN and Alternating
Decision Trees - ADTrees. The researches usually use WEKA software package for applying the algorithm
and obtaining results.
Keywords
Machine Learning, Autism Spectrum Disorder, diagnosis, features, ANN, ADTree,
WEKA.
WEKA.
Identifier
2637-2835
Publisher
International Burch University, Sarajevo, Bosnia and Herzegovina
Source
Journal of Natural Sciences and Engineering
Date
January, 2020