Predictive Demand Service for Public Transit Using CNN/ Atendimento Preditivo De Demanda Do Transporte Público Coletivo Usando CNN

Abstract

Several cities in Brazil undergo a territorial expansion and inhabitants constantly, this process is called urbanization. An uncontrolled urbanization generates many difficulties, highlighting the mobility of public transport, since many citizens depend on this mobility, we have, for example, public transport in Goiânia, which directly affects the living conditions of passengers. For your foreknowledge, a model capable of mirroring the performance of your demand is essential, providing that the system meets users in an acceptable way. A two-dimensional CNN is a CNN model that has a hidden convolutional layer that operates on a 1D sequence, it is a convenient mechanism to simulate a univariate forecast of time series of the predictive service of Goiânia's public transport. The method is equivalent to an analysis of the focal parts that make up the public transport system and how to represent it in the 1D convolutional