Prevalence and Characteristics of Children with Autism-Spectrum Disorders based on Hybrid RBF-ELMAN Approach

Early identification of autism spectrum disorder has long been a major challenge for the medical and health community. The key to good care and prompt intervention is an early diagnosis. It is crucial to identify the incidence and features of children with ASD, as research has shown that they have different degrees of difficulty with handwriting activities. In light of these difficulties, the authors of this study recommend creating a handwritten dataset for neural network classification that includes samples from people with and without ASD. Before extracting features using FDA and LDA, the preprocessing stages include motion correction, normalising intensity, spatial smoothing, data extraction, and time correction. The RBF-ELMAN algorithm is used to train the classification model. The suggested method outperforms more traditional methods in terms of classification accuracy, including ELM and RBF. These results highlight the promise of neural network classification based on handwriting for the diagnosis of autism spectrum disorder in youngsters. Improved early diagnosis and intervention tactics lead to better developmental outcomes for children with ASD. This work uses advanced preprocessing techniques and machine learning models to achieve this goal.