Machine Learning Application in the Field of Identification and Diagnosis of Autism Spectrum Disorder Based on Visualization Analysis

The traditional early screening and diagnosis of autism has certain subjectivity in the assessment methods, and with the rapid development of machine learning, it is gradually possible to use intelligent methods for large-scale senseless identification and diagnosis of autism. In this paper, we use citespace to visualize and analyze 458 sci literature on machine learning for autism identification and diagnosis from 2011 to 2023 in the core library of Web of Science, to understand the research hotspots and trends in this field, and the results show that the research hotspots focus on the use of brain imaging technology to study the neurological mechanisms of early stage patients with autism. The results show that the research hotspots focus on the use of brain imaging technology to study the neural mechanisms of early stage patients with autism spectrum disorders, etc., to extract the potential physiological markers, which are used as the input features of the machine learning technology to construct the autism identification and diagnosis model. In the future, researchers need to further innovate the machine learning algorithms and study deep learning algorithms with better performance to build more accurate models to provide objective evidence for early diagnosis.