This AI spots local COVID-19 outbreaks before they’ve spun out of control


Researchers have developed a new machine learning-based tool that detects potential COVID-19 outbreaks in individual US counties. The system makes its assessments by analyzing reported COVID-19 cases and deaths, rates of positive tests, face mask rules, social distancing policies, and changes in testing. It also incorporates the Centers for Disease Control and Prevention’s Social Vulnerability Index, which assesses how individual communities are prepared for hazardous events based on factors including poverty and population density. It then predicts how fast the virus is spreading by estimating how many days it will take for the number of cases to double. The forecasts are displayed on an interactive heat…

This story continues at The Next Web