Abstract
The data gathered from smart cities can help citizens and city manager planners know where and when they should be aware of the repercussions regarding events happening in different parts of the city. Most of the smart city data analysis solutions are focused on the events and occurrences of the city as a whole, making it difficult to discern the exact place and time of the consequences of a particular event. We propose a novel method to model the events in a city in space and time. We apply our methodology for vehicular traffic data basing our models in (convolutional) neuronal networks.