Predictive Model for Safe Unconfinement Commercial Environments in Times of COVID-19

Abstract

The purpose of the research is to reduce the chances of contagion by COVID-19 caused by lack
of confinement at work, through the implementation of a business environment management system with the
generation of indicators that facilitate decision-making for consumers. A predictive model based on
supervised neural networks trained with the linear regression algorithm will be used. The estimated
prediction results of the implemented model allow knowing the attention time associated with the studied
dates, obtaining optimized times of 15 to 18 minutes. It is concluded that the linear regression algorithm
offers a percentage of permissible error, this being 0.48%. The indicators generated for the estimated time of
care are close to reality, and the probability of contagion by COVID-19 is reduced.