• 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.104.2021.tde-27012022-105655
Document
Author
Full name
Juliana Shibaki Camargo
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2021
Supervisor
Committee
Diniz, Carlos Alberto Ribeiro (President)
Fiorucci, José Augusto
Souza, Anderson Luiz Ara
Title in Portuguese
Método bagging para aprimoramento de previsões de séries temporais
Keywords in Portuguese
Bagging
Bootstrap
Moving block bootstrap
Previsão
Séries temporais
Abstract in Portuguese
Diferentes metodologias são propostas e exploradas com o intuito de reduzir o erro de previsão de séries temporais. Uma estratégia que vem se apresentando bastante promissora consiste em combinar diferentes previsões de diferentes modelos a fim de se obter uma melhor acurácia, ou seja, um menor erro de previsão. Este trabalho teve como objetivo realizar um estudo e aplicação do método bootstrap aggregating, mais conhecido como bagging, para aprimorar previsões de séries temporais. Primeiramente, cada série temporal foi separada em série de treinamento e série de teste, e então utilizou-se a metodologia moving block bootstrap aplicada à série de treinamento para gerar diferentes séries reamostradas, realizar a previsão de cada uma delas e combiná-las, obtendo-se assim uma previsão final combinada. Posteriormente, a série de teste foi utilizada para calcular a acurácia dos modelos, individual e combinado. Foram realizados um estudo com séries simuladas e uma aplicação com séries temporais reais mensais. O modelo escolhido e ajustado para cada uma das séries foi obtido através da função auto.arima(), disponibilizada pelo pacote forecast do software R. As medidas de acurácia utilizadas foram o erro quadrático médio e sua raiz, o erro percentual absoluto médio arcotangente e o erro percentual absoluto médio simétrico. Ao final do estudo, explorou-se o impacto que a variação dos parâmetros da reamostragem do modelo combinado causa na previsão e foram realizadas comparações entre os métodos de previsão combinado e individual.
Title in English
Bagging method for improving time series forecasts
Keywords in English
Bagging
Bootstrap
Forecast
Moving Block bootstrap
Time series
Abstract in English
Different methodologies are proposed and explored aiming to reduce time series forecasting error. A promising approach consists in combining different forecasts from different models in order to get a better accuracy, i.e., a smaller forecast error. This work aims to review and apply the bootstrap aggregating method, also known as bagging, in order to improve time series forecasting. First, each time series is divided into training and testing time series, and then the moving block bootstrap methodology is applied to the training series to generate different resampled time series, and then forecasting for each one of the series is performed and combined, thus obtaining the final combined forecast. The test data set is used to calculate the accuracy of the models, individual and combined. A simulation study of time series and application to a real time series data sets are presented. The chosen and fitted model for each of the time series was obtained by using the function auto.arima(), available from forecast package, from R software. The accuracy measurements used were the mean square error and its root, mean arctangent absolute percentage error and the symmetric mean absolute percentage error. Finally, the impact on the forecasts of the combined model by varying the resampling method parameters was explored and comparisons between the combined and individual forecasting methods were also carried out.
 
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
2022-01-27
 
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 © 2001-2024. All rights reserved.