A Survey on Exploring Early Signs of Autism and User Perspectives

Abstract: Diagnosing Autism Spectrum Disorder (ASD) present challenges due to the absence of specific diagnostic tools for definitive identification. ASD is characterized by social communication challenges and restricted or repetitive behaviors, diagnosed based on observed behavior. This survey investigates the landscape of computational models for ASD detection, reviewing methodologies like Naïve Bayes, Support Vector Machine, Logistic Regression, KNN, Neural Network, and Convolutional Neural Network (CNN). The study extends its focus to include the development of an application to identify autism and support for parents. It addresses the increasing role of ML in medical research and aims to comprehensively explore the evolving landscape of ASD detection models. The paper delves into the intricacies of ASD diagnosis, emphasizing the potential synergies between ML methods and creative applications that enhance recognition and parental involvement.

Keywords: Autism Spectrum Detection (ASD), Machine Learning, Diagnostic Challenges, ML-Based Diagnosis, Supportive App.