• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.43.2020.tde-21122020-120638
Document
Author
Full name
Natália Villa Nova Rodrigues
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2020
Supervisor
Committee
Abramo, Luis Raul Weber (President)
Marra, Valerio
Sodre Junior, Laerte
Title in Portuguese
Classificação de quasares, estrelas e galáxias com técnicas de aprendizagem automática
Keywords in Portuguese
Estruturas em Largas Escalas
Levantamentos Fotométricos, Aprendizagem Automática.
Quasares
Abstract in Portuguese
A próxima geração de levantamentos astrofsicos contará com grandes quantidades de dados. Esse cenário motiva o uso de ferramentas de aprendizagem automática para classificar objetos observados como fontes pontuais de emissão. A seleção de quasares, em particular, é de fundamental importância para obter vnculos de parâmetros cosmológicos, investigar a evolução do universo e desvendar o mistério da energia escura. Neste tra- balho utilizamos algoritmos de aprendizagem automática para classificar quasares entre estrelas e galáxias. Em particular, desenvolvemos uma técnica para incluir as incertezas das medidas nesses algoritmos e mostramos, a partir de um modelo simplificado, que essa abordagem melhora a performance dos classificadores. Essas técnicas foram aplicadas aos dados de dois levantamentos fotométricos, S-PLUS e miniJPAS, que são caracteri- zados principalmente por suas configurações de filtros de bandas estreitas. As técnicas desenvolvidas aqui serão posteriormente utilizadas para construir catálogos de quasares e mapas de estruturas em grandes escalas.
Title in English
Classifying quasars, stars and galaxies with machine learning
Keywords in English
Large-Scale Structure
Machine Learning.
Photometric Surveys
Quasars
Abstract in English
The next generation of astrophysical surveys will rely on large amounts of data. This scenario motivates the application of machine learning tools to classify objects which are detected as point-like sources. The selection of quasars, in particular, is of fundamental importance to constrain cosmological parameters, to investigate the evolution of the uni- verse, and to unveil the mystery of dark energy. In this work we used machine learning algorithms to classify quasars, stars and galaxies. In particular, we developed a technique to include the uncertainties of the measurements in these algorithms and we proved, using a toy model, that this approach improves the performance of the classifiers. These tech- niques were applied in data from two photometric surveys, S-PLUS and miniJPAS, which are characterized mainly by their narrow-band filters. These techniques will be used to build quasar catalogs and maps of the large scale structures.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2021-01-22
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2021. All rights reserved.