Please use this identifier to cite or link to this item:
Title: Estudio comparativo de modelos de predicción estocásticos y heurísticos aplicados a la estimación de la calidad del aire
Author: Sánchez Pozo, Nadia Nathaly
Director: Solé Ribalta, Albert
Tutor: Trilles Oliver, Sergi
Keywords: prediction
air quality
Issue Date: Jul-2020
Publisher: Universitat Oberta de Catalunya (UOC)
Abstract: This work presents a comparative analysis of predictive models, applied to the estimation of air quality, currently among the world concerns is the concern about air pollution, therefore, in cities like London there are air pollution monitoring systems. The notable deterioration of air quality in London is an increasingly serious problem, considering that there is a direct relationship with respiratory and cardiac health problems being already the cause of death in that city. The objective of this study is to analyze and compare different predictive models, for determining which of them allows us to do a better prediction of London's air quality. To do so, an open data set recovered from the London Datastore portal is used, which are historical data corresponding to measurements of the city's pollutant monitoring system. These data are used to train the ARIMA, SVM, Neural Networks and Facebook Prophet algorithms. From the generated models generated, the ones reaching greater accuracy when predicting the concentration of air pollutants are to be determined.
Language: Spanish
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
nsanchezpoTFM0620memoria.pdfMemoria el TFM1,83 MBAdobe PDFThumbnail