Identification of Autism Spectrum Disorder (ASD) in Adults through Various Machine Learning Algorithms

This research study presents a comprehensive analysis of Autistic Spectrum Disorder (ASD) in adults, addressing its prevalence, diagnostic challenges, and the diverse range of support and intervention strategies available. The study analyzes the growing recognition of ASD as a lifelong condition and explores the unique social, emotional, and cognitive needs of adults on the spectrum. Our study also considers the value of early detection and treatment in enhancing the quality of life for adults with Autism Spectrum Disorder (ASD). Datasets with behavioral aspects are now more important than ever because of the global increase of ASD incidence. However, the lack of such datasets makes it difficult to perform in-depth analysis in order to improve the effectiveness, sensitivity, specificity, and prediction accuracy of ASD screening processes. Clinical or screening-oriented datasets for autism are now noticeably lacking. To fill this gap, this study proposes a new dataset that includes 20 unique traits, which was created exclusively for screening ASD in adults. This dataset includes ten behavioral characteristics assessed using the AQ-10-Adult and an additional ten personal characteristics that, according to behavioral science specialists, have shown useful in accurately differentiating people with ASD from control participants. This dataset shows potential for additional study, particularly in identifying important autistic characteristics and improving patient classification for ASD.