Time Series Forecasting for Improving Quality of Life and Ecosystem Services in Smart Cities

Quality of life is one of the factors that most influence the mood of citizens. As many studies have shown, one of the ways to increase the perception of quality of life are the actions on the Green Infrastructure of cities. Some studies have resorted to LSTM and ARIMA networks to make environmental predictions, however, as will be shown in this article, the seasonality of these models is a brake on the predictions. In order to perform efficient actions, an application case is presented, which has made use of cutting-edge methodologies thanks to IoT technology, Big Data and Artificial Intelligence to collect environmental data in order to perform time series prediction processes with them using GAM models, which have proven to be the most efficient during the tests carried out. Thanks to this work, it has been possible to obtain information on future environmental scenarios in order to make the best decisions on the influence that urban actions implemented by local authorities will have on citizens.