Back

Mathematical framework to model Covid-19 daily deaths

Barman, P.; Deb, N.; Mukherjee, S.

2020-05-22 epidemiology
10.1101/2020.05.18.20106104 medRxiv
Show abstract

The novel coronavirus (2019-nCoV) pandemic has caused widespread socio-economic disruption and, as of 04/07/2020, resulted in more than 72,614 confirmed deaths worldwide. Robust prediction of the trajectory of the death incidence curve is helpful for policy decisions during this ongoing crisis. We propose a non-parametric model to fit the number of daily deaths in a region, which hypothesizes that the death incidence curve will have a convex shape in the beginning, a concave shape near the peak, and a convex shape in the final stage of the death incidence curve after the peak. Using this, we performed robust short-term predictions on phases in five countries worldwide and five US states. Our analysis shows while the five states are all at peaks or past their peaks, US as a country is possibly not at peak yet. Our model can be easily fitted on daily death data from any region.

Matching journals

The top 5 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.