Abstract
Autism spectrum disorder (ASD) is a challenging developmental condition that involves restricted and/or repetitive behaviors and persistent challenges in social interaction and speech and nonverbal communication. There is not a standard medical test used to diagnose ASD; therefore, diagnosis is made by looking at the child's developmental history and behavior. In recent years, due to the increase in diagnosed cases of ASD, researchers proposed software-based tools to aid in and expedite the diagnosis. Considering the fact that most of these tools rely on the use of classifiers, in study, random forest, decision tree, k-nearest neighbors, and zero rule algorithms are used as classifiers, and their performances are compared using well-known performance metrics. As proven in the study, random forest algorithm can provide higher accuracy than the others in the classification of ASD and can be integrated into a computer- or humanoid-robot-based system for automated prescreening and diagnosis of ASD in preschool children groups.
TopBackground
In recent years, awareness of ASD has grown considerably, thereby resulting in a significant increase in the number of autism tests and diagnoses. Early diagnosis, even late diagnosis in some cases, can offer many benefits and notable relief. Autism screening tests can be held by healthcare professionals or professionals at a school setting or in the community. In these tests, with the help of a series of questions replied by parents and caregivers, professionals try to determine if children have developmental delays or are learning basic skills at the appropriate age. If needed, professionals play and talk with the children to see if they have any problems with movement, speech, and behaviors. On the other hand, screening tests just screen for characteristics of ASD and the screening results are just indicators, not diagnosis. A proper and correct diagnosis can only be made by healthcare professionals.
Key Terms in this Chapter
Receiver Operating Characteristic: A graphical plot that illustrates the diagnostic ability of a binary classification model as its discrimination threshold is varied.
Classification Accuracy: It shows the rate of correct classifications.
Autism Spectrum Disorder: A developmental disorder that affects social communication skills and behavior.
Cross Validation: It is a resampling procedure used to evaluate predictive models on a limited dataset.
Restricted and Repetitive Behaviors: They are the most common symptoms of autism spectrum disorder and involve ritualistic behavior such as rocking back and forth.
Precision-Recall Curve: It shows the trade-off between precision and recall for different thresholds.
Kappa Coefficient: A statistic that is used to measure inter-rater reliability for qualitative items.
Matthews Correlation Coefficient: A metric that shows the quality of classifications.
Evaluation Metric: It quantifies the performance of a predictive model.
Classification Algorithms: Algorithms used to map a given input data to a specific category.