Currently, most online apps displaying information about the COVID outbreak only show past and current data, but rarely give any insights on future development. With these limited ressources it is difficult for the governments and businesses to make informed decisions. It is also not easy to see if the taken meassurements are effective until a few days or even weeks later. Our goal was to create a simple and intuitive webinterface, that allows the user to simulate the development of the disease over a timeframe of up to three weeks. With this, the impact of different restrictions can be simulated and evaluated.
What it does
The app takes data specific to the behaviour of swiss citilziens, gathered by the Federal Statistical Office, to create an agent-based environment. An agent is a single individual in a closed population which interacts with other individuals. With these basics, the outbreak and development of a disease can be simulated. The used algorithm was developed by Ajelli, M., Gonçalves, B., Balcan, D. et al. in 2010. A presentation containing more detailed information about the project can be found here
What's next for COVID Predictor
Further data about population behavior and random social interaction can be added to improve the quality of the prediction as well as different forms of restrictions such as enforcing home-office or shutdown of educational institutions, etc.