Aggression in Children Predicted Using Machine Learning

A study led by Northeastern University professor Matthew Goodwin demonstrates that biosensor data and machine learning can predict aggressive behavior by youth with profound autism 3 minutes in advance with 80% accuracy. Biosensors can detect physiological changes indicating potential aggression, giving caregivers time to intervene and prevent volatile situations.

The study, funded by the Simons Foundation, Nancy Lurie Marks Family Foundation, and the Department of Defense, utilized wristwatch-like biosensors on youth with profound autism in psychiatric hospitals, collecting data on 6,665 aggressive behaviors. The findings could lead to personalized intervention plans and reduced reliance on emergency services for families of individuals with profound autism