A Novel Monte Carlo Simulation of the Spread of Coronavirus

Inspiration

The current coronavirus epidemic is a global concern and has greatly altered the quality of living in China and it neighboring countries. We joined this hackathon in hopes to provide a step towards a solution.

What it does

The model simulates the number of infected civilians, the number of deaths, and the number of recovered civilians given information on the hospital housing/bed limit, quarantine stringency, recovery days, as well as other factors over a given n window of days.

How I built it

We used a .csv dataset collected by doctors at Johns Hopkins University to model a normal distribution of mortality/recovery rates of coronavirus; we then used this information alongside the infection rates published by the WHO to create a mathematical model of citizens that could be further affected by those with the virus.

Challenges we ran into

While coronavirus is a disease spread by person to person contact, the data representing its transmission was at a population level. Furthermore, the estimated recovery time of an infected patient from coronavirus is documented >= 21 days; therefore, we needed to store an array of infected civilians and the day in which they were affected. Finally, it took us a long time to figure out the hyperparameters that most appropriately modeled the spread of coronavirus, especially since there were not much information on the disease in its early days.

Accomplishments that we're proud of

We are quite proud at how accurately/closely it modeled the spread of coronavirus in both the United States and in Wuhan, China. Furthermore, we are proud of the social impact that such a model can potentially tell us regarding the magnitude of medical resources that countries need to allocate.

What we learned

We learned that modelling the transmission of disease from city to city is a much more complex issue than we imagined. We also learned much more about coronavirus and how it spreads, and how a quarantine may be extremely affective in combating the issue.

What's next for A Novel Monte Carlo Simulation of the Spread of Coronavirus

Note that this model can be used for other diseases; we are interested in generalizing the project

Built With

• jupyter
• python