The population in large urban centers has increased significantly over the last decades, resulting in a multitude of economic, social and environmental problems, putting their sustainability at risk. Therefore, the need for new ideas based on the saving and proper management of natural resources and available capital is imperative, focusing on improving the living conditions of present city residents without jeopardizing the ability of future generations to cover their own Needs. Here comes the idea of the "smart city", whose implementation is drastically contributing to "Big Data" and Information and Communication Technologies (ICT).
With the help of ICT it is possible to exploit the enormous amount of data arising from the day-to-day activities of citizens and to develop numerous applications and models for the export of motifs and the identification of correlations between urban data in order to anticipate critical situations such as extreme weather, congestion and high levels of air pollution, and thus to prevent the negative effects they cause. In the context of this diploma thesis, using data as a case study of the city of Dublin, data derived from sensors and traffic flow monitoring stations, weather conditions and emissions of gaseous pollutants were explored and possible interactions / interrelationships were investigated. In particular, data conversion and processing was performed using Python and its appropriate libraries (NumPy and Pandas), for the statistical analysis and visualization, the matplotlib and seaborn libraries were used, while for the development of the prediction model the scikit-learn library.
The results of the analysis and the regression model that emerged, if supported by the proposed modifications and extensions, can help to adopt modern techniques to locate existing urban problems and to promote development plans and financial tools to deal effectively with them.