In August this year, Google, in collaboration with the Harvard Institute of global health, launched the covid-19 public prediction model, which provides new crown cases, deaths, ventilator availability, ICU utilization and other indicators for all States and counties in the United States. Today, Google released an improved new crown prediction model, which increases the time span and adds new forecast areas. The new model was trained using public data from Descartes laboratories, the U.S. Census Bureau, Johns Hopkins University and elsewhere. < / P > < p > Google said the new outbreak public forecast aims to provide quick response resources for the public sector, Healthcare Architecture and other affected organizations. The model allows us to conduct new crown tests and public health interventions on a county by county basis, with the strongest ability to respond quickly to the development of the epidemic. With the help of this model, state and county health departments can use Infection Prediction to provide information for detection strategies and identify areas with outbreak risk. Medical institutions can also use this model to incorporate the predicted number of cases as data points into the resource planning of PPE, staffing and scheduling. < p > < p > at first, Google’s new crown epidemic public prediction model can predict the region in the next 14 days. Through a long time of large amount of data learning, the accuracy rate has been improved by about 50%, and the situation within 28 days can be predicted. And it’s got public forecasts for other countries that it’s already rolling out. Like the United States, the forecast is based on public data such as the official Japanese epidemic situation report. After training, it can predict the number of confirmed cases, the number of hospitalized people, the number of deaths and the number of rehabilitated patients every day, and predict the new crown epidemic situation in Japan’s counties in the next 28 days. < / P > < p > Google said that the original forecasting model was improved and could be customized according to data sets and new situations and new problems. The company is also developing an AI driven & quot; what if & quot; model to provide decision-making for new outbreaks and other infectious diseases. Privacy Policy