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

where p i are the mixture probabilities, η i are the dispersion parameters of the negative-binomial distributions and Z i are the competing predictors obtained from mortality curves in ahead-in-time countries i = 1, . . . , n. Predictor Z i is built from a smoothed, scaled and delayed version of the process Y i , which is the analogue of Y 0 for country i. The smoothing aims at mitigating events that are specific to country i. The scaling aims at homogenizing population sizes with the focal country 0. The delay, denoted by δ i , accounts for temporal lag in the start and progression of the epidemics between country 0 and i. This delay is calculated as the duration between the date τ of the last observation used in the analysis and the (anterior) date at which the smoothed version of Y i reached the value Y 0 (τ ).