Machine learning Approaches for Autism Spectrum Disorder Prediction

The neurodevelopmental illness known as autism spectrum disorder is a lifelong condition marked by a persistent or limited development of social-communication skills, cognitive abilities, activities, and behavior. The symptoms of autism are easier to see and identify in children between the ages of two and three. A prediction system based on traditional machine learning techniques, such as support vector machines, random forests, multiple layer perceptrons, naive Bayes, convolution neural networks, and deep neural networks, accounts for most of the research on autism spectrum disorder. Accuracy, precision, and recall are among the performance indicators used to verify the proposed models. This study investigated and evaluated the prediction of autism spectrum disorders using standard criteria including application type, simulation method, comparison approach, and input data. The purpose of this work is to offer a consolidated framework for autism spectrum disorder prediction to researchers. With the best results, the random forest method performed better than traditional machine learning algorithms. 97.8% accuracy percentage was achieved. Readers can better comprehend the underlying mechanisms and structures of the investigated frameworks with the aid of these work flow representations.