• 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.45.2020.tde-10122020-165920
Document
Author
Full name
Thiago Ildeu Albuquerque Lira
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2020
Supervisor
Committee
Finger, Marcelo (President)
John, Vanderley Moacyr
Lotufo, Roberto de Alencar
Title in Portuguese
Modelos neurais para regressão de séries temporais
Keywords in Portuguese
Deep learning
Inferência Bayesiana
Séries temporais
Abstract in Portuguese
Dada a crescente importância da predição de resistência compressiva do cimento para um uso mais eciente de recursos na indústria, literatura recente busca analisar quais modelos estatísticos podem auxiliar o processo indústrial. Esse trabalho documenta a aplicação de técnicas de Deep Learning Bayesiano para a geração de predições temporais robustas e conáveis para a resistência compressiva do cimento. Os resultados mostram que técnicas de Inferência Bayesiana para modelos de Aprendizado Profundo promovem um ganho sensível de acurácia para o problema de predição de RC, com o benefício adicional das características probabilísticas das predições, tornando-as mais seguras para o possível uso no chão de fábrica.
Title in English
Neural models for regression on time series
Keywords in English
Bayesian inference
Deep learning
Time series
Abstract in English
Given the increasing importance of the prediction of the cement compressive strength for a more ecient use of resources on the industry, recent literature has been experimenting with statistical models to aid the industrial process. This thesis studies the application of Bayesian Deep Learning techniques to achieve robust and accurate predictions of the compressive strength. The results show that Bayesian Inference techniques applied to Deep Learning models promote a sensible increase of accuracy for the problem of CS predition, with the additional benet gained from the probabilistic nature of the predictions, making them more suitable to be used on the factory oor.
 
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.
diss_corrigida.pdf (1.62 Mbytes)
Publishing Date
2020-12-18
 
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.