Back

Rule of thumb in human intelligence for assessing the COVID-19 outbreak in Japan

Kurita, J.; Sugawara, T.; Ohkusa, Y.

2021-01-26 public and global health
10.1101/2021.01.20.21250204 medRxiv
Show abstract

BackgroundThe COVID-19 outbreak in Japan exhibited its third peak at the end of 2020. Mathematical modelling and developed AI cannot explain several peaks in a single year. ObjectThis study was conducted to evaluate a rule of thumb for prediction from past wave experiences. MethodWe rescaled the number of newly infected patients as 100% at the peak and checked similarities among waves. Then we extrapolated the courses of the third and later waves. ResultsResults show some similarity around the second and the third waves. Based on this similarity, we expected the bottom of the third wave will show 2131 newly positive patients including asymptomatic patients at around the end of February, 2021. Discussion and ConclusionWe can infer the course of the third wave from similarity with the second wave. Mathematical modelling has been unable to do it, even when AI was used for prediction. Performance of the rule of thumb used with human intelligence might be superior to that of AI under these circumstances.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.