Tasa generadora de viajes para el puerto de Montevideo. Una propuesta metodológica

The predictive analysis is becoming increasingly important in the definition of transport policies and in the design of operational activities. In the present article the developed methodology in order to attain an accurate estimation of the movement of containerized cargo on the entrance to the port of Montevideo is presented. Over the last years, the port of Montevideo has gone through an important growth. This, together with its operative features, its placing and interrelation with the country, and its development outlook for the future, requires the use of modern management tools, providing an important opportunity for investigation. The bibliography concerning the modeling of incoming/outgoing containerized cargo in ports is scarce and not always useful. In this article count models were developed, mainly Poisson regression, that explain the production/attraction of containerized cargo trips according to the previous declaration of wharf operation. This Poisson regression included binary and auto regressive lags variables in order to deal adequately with the temporary dimension of the problem.