Predicting car accidents in Barcelona using a Random Forest model

All of RECERCAT

To access the full text documents, please follow this link: http://hdl.handle.net/2117/100298

Title:	Predicting car accidents in Barcelona using a Random Forest model
Author:	Moreno Hoyo, Alberto
Other authors:	Universitat Politècnica de Catalunya. Departament d'Estadística i Investigació Operativa; Langohr, Klaus
Abstract:	The city council of Barcelona releases traffic accident data every year, making them public through its portal OpenDataBCN. This Thesis merges these data with a weather data set made available by Wunderground.com and with holiday data acquired through web scraping at DateAndTime.com, in order to extract insights about local traffic accidents and, ultimately, build a Random Forest model to estimate accident probabilities during a given time interval in any point of the city.
Subject(s):	-Àrees temàtiques de la UPC::Matemàtiques i estadística::Estadística matemàtica::Anàlisi multivariant -Multivariate analysis -Open data -Random forest -Accident -Traffic -Barcelona -Anàlisi multivariable -Classificació AMS::62 Statistics::62H Multivariate analysis
Rights:
Document type:	Research/Master Thesis
Published by:	Universitat Politècnica de Catalunya; Universitat de Barcelona
Share:

Coordination

Supporters