Detection of Autism Spectrum Disorder by Using Common Machine Learning Algorithms

Communication, social interaction, behavior, and sensory processing are all negatively impacted by the neurodevelopmental disease known as autism spectrum disorder (ASD). The word "spectrum" is included in the name because there is such a broad variety of symptoms and degrees of impairment. The health and well-being of children with autism spectrum disorder (ASD) can be greatly improved by early diagnosis and intervention. The current detection techniques are based on the opinion of specialists, which is subjective and expensive. In this research, we suggest a machine learning strategy for ASD diagnosis that incorporates a suite of machine learning techniques. Using it has the potential to save money and increase detection efficiency. To begin, we performed pre-processing to extract characteristics from the data. Then, classify the data with a 0.997% accuracy using four machine learning (ML)-based methods: The Random Forest, logistic regression, support vector machine (SVM) and K-nearest neighbors algorithm (KNN). The findings of this research provide promising evidence for the efficacy of the machine learning classification technique used to diagnose ASD at an early age.