Abstract
This work proposed predictive modeling approach for Autism Spectrum Disorder (ASD) assessment, employing machine learning algorithms to analyze multifactorial features. Utilizing a diverse dataset that includes demographic, behavioral, and clinical information, our models aim to predict the likelihood of ASD. The chosen evaluation metric is the Area Under the Receiver Operating Characteristic Curve (AUC-ROC Score), ensuring robust model performance assessment. Through rigorous experimentation, we demonstrate the effectiveness of our methodology in accurately identifying individuals at risk of ASD. This research contributes to advancing early detection methods and enhancing our understanding of the intricate interplay of features influencing ASD, laying the groundwork for more informed diagnostic strategies.