Regresar

Variational inference for the estimation of covid-19 mortality in portoviejo

Abstract:

The severe acute respiratory syndrome pandemic by SARS-CoV-2 is caused, millions of confirmed cases and a high number of deaths are reported, the population dynamics are altered, a strong socio-economic impact, the collapse in the health system, the collapse in the education system, unemployment, are caused, among others. In this work, a logistic regression model is proposed, the dynamics of deaths in the period of the pandemic is modeled, two computational algorithms IVDA and NUTS are implemented, the posterior distribution in high dimensions is generated. The interpretation of the parameters is important, that to the variable of deaths from COVID-19 or no COVID-19 are related. The algorithms through real data are illustrated, the proposed model with adequately and reliably of the dynamic of the randomness of the deceased is shown. To measure the relative success of the algorithms, three goodness-of-fit measures are estimated. Obtaining small errors.