A new study has found that machine learning can accurately predict autism spectrum disorder (ASD) using a minimal set of features from background and medical information. The most important predictors were found to be developmental milestones and eating behavior. This could potentially allow for early screening and diagnosis of ASD in children.
