The Covid-19 virus is the same all over the world. Still, different countries have different ways of dealing with the infections. Why is it that Italy has 10 times more cases than South Korea, even thought the pandemic started more or less at the same time in both countries? We figured that one of the reasons might be the measures taken by the government in order to fight the virus.
What it does
We chose seven reference countries to analyse: Switzerland, China (Hubei region mostly), Italy, South Korea, Spain, the USA and Singapore. We analysed different measures taken or recommended by the local government: Degree of lockdown, limit on the number of people in a group, schools being closed, stores being closed, working from home, travel restrictions, and number of tests done. These differ from country to country in the time at which they were put into action, as well as the strictness of their enforcement. The output is an optimal combination of such safety measures based on our data.
How we built it
We built a model using machine learning techniques to find out which of measures are the most effective and to what degree. We used python together with the libraries pandas and sklearn.
Challenges we ran into
-Due to insufficient documentation, gaining all the needed information about our reference regions was sometimes tricky.
-For some countries (such as the USA), the enforced safety measures vary tremendously from state to state, which made a wholistic analysis of the country's approach to health regulations quite difficult. Simplifications had to be made.
-Due to the 14-day incubation period of the virus and the fact that confirmed infections recently started to roll in and many are yet to come, there's a relatively long ramp-up phase underlying our datasets. This lead to our results being quite biased, since the rates in the early stages of our collected data are small.
Accomplishments that we're proud of
We managed to build a functioning model.
What we learned
We have gained valuable insights into data analytical approaches and now understand how the measures taken in a pandemic affect the spread of a virus. Our interpretation of the results is as follows: It seems as though a complete shutting-down of schools and shops together with a strict travel ban is the most effective combination of measures. Furthermore, performing a lot of tests is also useful. However, our results do NOT confirm the fact that complete quarantine and shutting-down of offices is ineffective. Our results should be handled with caution, since the information acquired about the countries were subject to simplifications and may not be reliable. In order to improve the scientific approach, countries should be more open to transparency when it comes to the safety measures they are taking
Try It out
pandas, python, sklearn