Wind pattern analysis applied Tokyo 2020 Olympic Game

Di Francesco, Fabio (2019) Wind pattern analysis applied Tokyo 2020 Olympic Game. [Laurea magistrale], Università di Bologna, Corso di Studio in Ingegneria informatica [LM-DM270]
Documenti full-text disponibili:
[img] Documento PDF (Thesis)
Disponibile con Licenza: Creative Commons: Attribuzione - Non commerciale - Non opere derivate 3.0 (CC BY-NC-ND 3.0)

Download (4MB)

Abstract

The following master thesis is the product of the work carried out during the Erasmus exchange of the year 2017-2018 that involved the author, exchange student from the University of Bologna, the Universitat Politècnica de Catalunya , TriM, an italian company with a strong knowledge of weather data and forecasting, and Meteocat, the public meteorological company of Catalonia in a collaboration aimed to find new methodologies for the processing of meteorological data. The reason that motivated this work is dictated by the increasing amount of weather data available today, that necessarily drives the weather forecasting in a more automated procedure that reduces the time needed to generate a forecast and the intervention of a human, in the figure of a meteorologist, in the analysis of the data. This allows to process more data and thus having predictions that take advantages of the usage of many information that could result in improved forecasting. The development in the field of machine learning allows today to treat a vast amount of information in an automatic way, leaving the analysis process to the machines, freeing the user of this time-consuming task. And unsupervised learning is the branch that can process data that are not labelled nor preprocessed, speeding up the data mining. The goal of this thesis is to apply unsupervised learning techniques to this scope, taking inspiration from the available literature that experimented in this field and combining different solutions into a new technique that aims to reduce the human decision in the process of the recognition of wind patterns and improve the automationof the whole process.

Abstract
Tipologia del documento
Tesi di laurea (Laurea magistrale)
Autore della tesi
Di Francesco, Fabio
Relatore della tesi
Correlatore della tesi
Scuola
Corso di studio
Ordinamento Cds
DM270
Parole chiave
Machine Learning,Data mining,Data anaylisis,Clustering
Data di discussione della Tesi
22 Luglio 2019
URI

Altri metadati

Statistica sui download

Gestione del documento: Visualizza il documento

^