Title: Early dynamics of COVID-19 in Algeria: a model-based study

Abstract: We propose a compartmental model based on genetic fitting algorithm to estimate early stage COVID-19 epidemic dynamics in Algeria. Provided reasonably accurate cumulative confirmed cases, recoveries and deaths numbers, and using cross-validation method to overcome the overfitting problem, our model has the capacity to estimate key epidemic parameters as reproduction number, latent time, infectious time, recovery and fatality rates. Moreover, we advance approximate nowcast and forecast of the non-measurable asymptotic infected portion of the susceptible population and the incidence of disease as well as quarantined active cases. Epidemic peak-time is estimated and CFR and IFR are calculated. Information provided by our model could help establish a realistic assessment of the situation for the time being and near future in order to apply the appropriate public health measures.