Por favor, use este identificador para citar o enlazar este ítem:
http://dspace.udla.edu.ec/handle/33000/9288| Tipo de material : | bachelorThesis |
| Título : | Modelo de predicción de la calidad del aire a partir de datos meteorológicos e información del tráfico automovilístico |
| Autor : | Ramírez Suárez, Jessie María |
| Tutor : | Rybarczyk, Yves Philippe |
| Palabras clave : | CALIDAD AMBIENTAL;CALIDAD DEL AIRE;DATOS ATMOSFÉRICOS;TÉCNICAS DE PREDICCIÓN;TRÁFICO URBANO |
| Fecha de publicación : | 2018 |
| Editorial : | Quito: Universidad de las Américas, 2018 |
| Citación : | Ramírez Suárez, J. M. (2018). Modelo de predicción de la calidad del aire a partir de datos meteorológicos e información del tráfico automovilístico (Tesis de pregrado). Universidad de las Américas, Quito. |
| Resumen : | La contaminación del aire representa un importante riesgo medioambiental para la salud. Únicamente buscando disminuir los niveles de esta contaminación, los países pueden reducir la carga de morbilidad derivada de accidentes cerebrovasculares, cánceres de pulmón y neumopatías crónicas y agudas, entre ellas el asma... |
| Descripción : | Air pollution poses a major environmental risk to health. Only by seeking to reduce the levels of this pollution, countries can reduce the burden of disease resulting from stroke, lung cancers and chronic and acute pneumopathies, including asthma. (World Health Organization, 2016). In the article Modeling PM2.5 Urban Pollution Using Machine Learning and Selected Meteorological Parameters published on June 18,2017 by Yves Rybarczyk, Mario Gonzalez and Rasa Zalakeiciute, professors at the University of the Americas; they propose an automatic learning approach based on six years of analysis of meteorological data and pollution in order to predict concentrations of fine particulate matter (PM2).5) from wind levels (speed and direction) and precipitation by applying the Linear Regression algorithm. In this thesis document, Machine Learning concepts are defined as: Supervised Learning, Classifiers, Regression, among other topics. In order to obtain the information, two applications were executed in data collection, which were obtained by the following methods: traffic time data managed by Google Maps. Screens captured from Google Maps, obtaining traffic information represented in red, orange and green, graphically rectangular shape. And finally, data was collected by making a circular area cut. All the information obtained in these three methods was stored in a CSV file, which is refreshed every 10 minutes. Once the CSV files had been obtained, the data preprocessing was carried out, cleaning the information by filling the empty spaces of the CSV files with the ? sign. To load them into the Weka software, we worked with Linear Regression, classifiers such as Neural Networks, SVM, Logistic Regression and KNN. Using the results obtained, confusion matrices were developed for the construction of the ROC curve in order to obtain the best performing supervised learning technique. In the future, an application can be developed with the aim of making this entity aware of the level of contamination to which the population is exposed and taking corrective measures. |
| URI : | http://dspace.udla.edu.ec/handle/33000/9288 |
| Aparece en las colecciones: | Ingeniería en Sistemas de Computación e Informática |
Ficheros en este ítem:
| Fichero | Descripción | Tamaño | Formato | |
|---|---|---|---|---|
| UDLA-EC-TIS-2018-02.pdf | 16,36 MB | Adobe PDF | Visualizar/Abrir |
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons
