Machine learning techniques will take an increasingly central role in the distributed sensing realm and specifically in smart cities scenarios. Pervasive air quality monitoring as one of the primary city service requires a significant effort in term of data processing for extracting the needed semantic value. In this paper, after briefly reviewing the emerging relevant literature, we compare several machine learning tools for the purpose of devising intelligent calibration components to be run on board or in cloud computing architectures for pollutant concentration estimation. Two cities field experiments provide the needed on field recorded datasets to validate the approaches. Results are discussed both in terms of performance and computational impact for the specific application. © Springer International Publishing AG 2017.
Computational intelligence for smart air quality monitors calibration
Fattoruso, G.;Salvato, M.;De Vito, S.;Esposito, E.
2017-01-01
Abstract
Machine learning techniques will take an increasingly central role in the distributed sensing realm and specifically in smart cities scenarios. Pervasive air quality monitoring as one of the primary city service requires a significant effort in term of data processing for extracting the needed semantic value. In this paper, after briefly reviewing the emerging relevant literature, we compare several machine learning tools for the purpose of devising intelligent calibration components to be run on board or in cloud computing architectures for pollutant concentration estimation. Two cities field experiments provide the needed on field recorded datasets to validate the approaches. Results are discussed both in terms of performance and computational impact for the specific application. © Springer International Publishing AG 2017.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
