Development of autism spectrum disorders classification model for toddlers using machine learning technique

Abstract

         A neurological condition known as an autism spectrum disorder (ASD) may affect a person's speech, cognitive abilities, language acquisition, and communication abilities. Understanding and interacting with others is challenging for people with ASDs. It is mainly triggered by genetics or outside influences; however, early diagnosis and treatment can improve outcomes. Currently, only clinically validated assessments are used to diagnose ASD, leading to longer diagnostic times and higher medical costs.  Machine learning is utilized to increase diagnosis accuracy and speed, supplementing traditional methods. In this study, we applied a machine learning approach using genetic programming (GP) to the TASD dataset to detect ASD cases. The suggested approach achieved an accuracy of 98.48%, which outperformed the state-of-the-art by 10%. These results from the proposed approach reflect its robustness and ability to identify ASD cases in the early stages with high accuracy.