A program to predict the number of deaths from the number of new coronavirus infections in Japan in 2020 Also, it is a program that simply analyzes the data of Tokyo.
It is simple, lacks test data, and the error is 59 people, so the accuracy is poor. However, the accuracy of the explanatory model and the objective variable itself is 90% or more, and the regression data are correlated, so this time it is a single regression of only the number of infected people, but probably work, diet, family composition etc. are added. I am. It will be helpful when doing multiple regressions at the event.
In the future, we plan to improve prediction accuracy and perform multiple regression analysis.
If you have an interesting pull request, such as a change, modification, or addition of data, please do so.