A Comparative Analysis: Autism Spectrum Disorder Among Children Using Classification Models

Autism is a neurological condition marked by challenges with communication and social engagement, some recurring interests, and habits. It is a very complex and diverse condition that affects individuals, and there is no prescribed treatment for autism. Machine learning models have such potential to understand autism and provide more effective interventions for autistic patients. However, learning models are particularly challenging to read, making it challenging to understand how the model makes decisions and identify any biases in the model. These are important in the context of research on autism, where it is possible to assess the effects of inaccurate diagnosis or improper treatment recommendations. In this work, various algorithms are applied to an autism spectrum disorder (ASD) dataset with 931 samples and 103 features. Experimental results show that the predictive model Random Forest gives the highest accuracy, precision, recall, and Fl-Score with 100% compared to LR, GNB, KNN, SVM, and ID3.