The current COVID-19 wave will likely be mitigated in the second-line European countries

Samuel Soubeyrand; Melina Ribaud; Virgile Baudrot; Denis Allard; Denys Pommeret; Lionel Roques

Second, we apply an estimation approach grounded on a weighted penalized likelihood to estimate p i , η i and δ i . Thus, we obtain a probabilistic assessment of which predicting trajectory is followed by the mortality curve of the focal country. In addition, we evaluate with the dispersion parameter η i how much the trajectory of the focal country is apart from the trajectory of the predicting country i. We also evaluate with the delay parameter δ i how much the focal country is late with respect to country i. This analysis can be applied on mortality curves observed up to the last observation date available in the data set. It can also be applied by ignoring, voluntarily, a part of the time series, to measure the stability of mortality progression in comparison with available predictors or, in the opposite way, to demonstrate changes in the epidemic regime. Applying the analysis back in the past also allows assessing forecast accuracy (like hindcasting in model-based climate-change studies).