CoViD19 Meta heuristic optimization based forecast method on time dependent bootstrapped data

Livio Fenga; Carlo Del Castello

is the (which was not peer-reviewed) The copyright holder for this preprint . https: //doi.org/10.1101/2020.04.02.20050153 doi: medRxiv preprint As an approximation, confidence bounds from normal distibution are derived for each of the parameters, such that with 0.01 significance level P r(119401 ≤ h ≤ 124955) = 0.99 P r(36.41 ≤ µ ≤ 37.07) = 0.99 P r(10.6 ≤ σ ≤ 10.9) = 0.99 µ confidence interval points out a peak day of daily cases between March 25 th and 26 th , while h magnitude parameter shows an asymptote of totale active cases curve between 120000 and 125000. The new cases curve has an asymptotic behavior, so cutting tail beyond a 0.1 cut-off for new infections, the pandemic time window is hypothetically over after May 16 th . This behavior is clearly described in the below charts built considering Gaussian and Cumulated Gaussian around 99% confidence lower and upper bound for each parameter.