In the era of big data, the ones extracted from mobile phones increase the potentiality for forecasting the amount of traffic flows in a specific area and in a specific time interval. Traffic flows among two regions, however, present a peculiar time series structure, where the daily and the weekly periods are strongly pronounced. For a good prediction performance, one needs to consider a time series modelling structure that takes into account for this kind of complex seasonality. Using one year of mobile phone traffic flows data, retrieved at one hour intervals, in this short paper we aim at forecasting with Harmonic Dynamic Regression models the flows in the strongly urbanized and flooding risk area of the Mandolossa, at the western outskirt of Brescia.
Big data estratti dai telefoni cellulari incrementano le potenzialità nella previsione dei flussi di traffico in una determinata area e in un determinato intervallo di tempo. I flussi di traffico presentano una peculiare struttura di serie storica, dove le periodicità giornaliere e settimanali sono fortemente pronunciate. Per una buona ”performance” previsiva, è necessario quindi considerare dei modelli che tengano conto di questo tipo di stagionalità complessa. Utilizzando un anno di dati sui flussi di traffico di telefonia mobile, recuperati ad intervalli di un’ora, in questo articolo ci poniamo l’obiettivo di prevedere, con un modello appropriato, i flussi nell’area urbanizzata e a rischio di alluvione della Mandolossa, alla periferia occidentale di Brescia.
(2022). Forecasting Traffic Flows with Complex Seasonality using Mobile Phone Data . Retrieved from http://hdl.handle.net/10446/227962
Forecasting Traffic Flows with Complex Seasonality using Mobile Phone Data
Metulini, Rodolfo;
2022-01-01
Abstract
In the era of big data, the ones extracted from mobile phones increase the potentiality for forecasting the amount of traffic flows in a specific area and in a specific time interval. Traffic flows among two regions, however, present a peculiar time series structure, where the daily and the weekly periods are strongly pronounced. For a good prediction performance, one needs to consider a time series modelling structure that takes into account for this kind of complex seasonality. Using one year of mobile phone traffic flows data, retrieved at one hour intervals, in this short paper we aim at forecasting with Harmonic Dynamic Regression models the flows in the strongly urbanized and flooding risk area of the Mandolossa, at the western outskirt of Brescia.File | Dimensione del file | Formato | |
---|---|---|---|
IES proc 2022_OD_flows.pdf
Solo gestori di archivio
Versione:
publisher's version - versione editoriale
Licenza:
Licenza default Aisberg
Dimensione del file
2.22 MB
Formato
Adobe PDF
|
2.22 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
Aisberg ©2008 Servizi bibliotecari, Università degli studi di Bergamo | Terms of use/Condizioni di utilizzo