Optimized Machine Learning And Deep Learning Classification Approaches For Early Detection Of Autism Spectrum Disorder

Abstract

Machine learning is now one of the most important technologies. Many academic disciplines, but especially data analytics, have benefited greatly from machine learning. Machine learning has produced major improvements in healthcare, in part because of the enormous volumes of data being collected and examined. Improvements in diagnosis have been made possible by new technologies that have matured over the last decade and are now commonplace in all medical facilities. The purpose of this proposed research work is to investigate and implement a wide variety of machine learning and deep learning algorithms for early prediction of ASD in very large datasets. Disorders on the autism spectrum are neurological and developmental in nature. Different people with autism will have varied degrees of difficulty with social interaction, conduct, language (spoken and non-spoken), and other aspects of their own unique identities. Possible autism onset ages range from 18 months to 3 years. Autism has been linked to both hereditary and environmental causes, although none has been singled out as the only cause. newlineThe accuracy of ASD screening results is highly dependent on the expertise of the evaluator. The standard way of detecting autism spectrum disorder is a manual approach that screens the youngsters and makes a diagnosis of ASD if they fall within the specified score scale. Predicting whether or not a kid will develop autism and the degree to which they will be affected is not straightforward; even the most experienced clinicians have challenges and ambiguity in this regard. Parents usually want to see quick results. Problems like ASD diagnosis and skill level prediction may be solved with the help of Machine Learning and Deep Learning methods. In this respect, this proposed research work will aid doctors in determining the best course of action. newline newlineThe goal of this proposed research work is to demonstrate how to train a data model on one set of data and then test and assess it on another.