LWTSEO-VT: Lightweight TwinScope Election Optimizer based-Vision Transformer for Autism Spectral Disorder Detection for Toddlers and Adults

autism spectrum disorder (ASD) is a neurological condition that disturbs an individual's capability to attach and communicate with others. It instigates in childhood and continues beyond adolescence and adulthood. Because of complicated procedures and limitations of existing methods, it has been difficult to diagnose. One unsolved problem is determining how to learn discriminative deep representations for building computational models. In this paper, a new Lightweight TwinScope Election Optimizer based-Vision Transformer (LWTSEO-VT) for ASD recognition is proposed for toddlers and adults. First, pre-processing is applied to improve the data quality and then extracted the features using Path-selective LASSO regression model. Lastly, a Lightweight TwinScope Vision Transformer (LWTS-VT) based on election optimizer is used for detecting autism. The LWTSEO-VT model implemented using python platform using autism screening data for toddlers and autism screening data for adults datasets. The efficacy and dependability of the LWTSEO-VT model have been demonstrated by quantitative comparisons