This application is the result of a joint work by professors and graduate students from the Statistics Department at UFMG. It originated as a challenge in a graduate course after the suspension of classes due to Covid19.
In the current configuration, the application has two main types of results: short-term and long-term predictions. The first refers to prediction of number of deaths and confirmed cases for the immediate future (up to 1 to 2 weeks). The second type of forecast is more comprehensive and aims to provide a full picture of the pandemic: when will the number of cases stop growing and start to decline? How many people will get sick? When can we expect the pandemic to end?
Our forecasts are updated daily, and may change based on new data that arrives every day. The predictions are accompanied by the respective probability intervals (or credibility, in statistical jargon) so that the user is always aware of the true uncertainty associated with each forecast provided. Another important point is that the predictions are always based on the maintenance of conditions on the day the forecast was made, including isolation conditions. Changes on these conditions can cause substantial alterations on the predictions.
For more information on the statistical model behind the forecasts, see the Methodology section. The events that team members have participated in and the news mentioning our project are in the Events section. Follow our latest posts and news on our Blog. All the source code used in this project is available in our GitHub repository.
The application with short and long term forecasts can be accessed in the Predictions section. In the Instructions section, you can follow a detailed description of all the features of our application.