• 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
10.11606/D.55.2018.tde-12032018-105302
Document
Author
Full name
Vanda Donizetti Redondo Silveira
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 1999
Supervisor
Committee
Rodrigues, Josemar (President)
Andrade Filho, Marinho Gomes de
Leite, Jose Galvao
Title in Portuguese
Inferência Bayesiana para pesquisa de mercado com erros de resposta utilizando modelos mistos
Keywords in Portuguese
Não disponível
Abstract in Portuguese
Inferência Bayesiana para pesquisa de mercado incluindo erros de resposta é estudado como uma mistura de duas distribuições de Bernoulli. Como a análise Bayesiana geralmente implica em cálculos complexos, o método de Monte Cano com dados ampliados é desenvolvido para obter os resumos marginais a posteriori. Variáveis latentes foram introduzidas para indicar qual componente da mistura gerou a informação com erro de classificação. Também, um procedimento Bayesiano baseado no conceito de "p-value" e na distância de variação total foi introduzida para medir o efeito do erro na distribuição marginal a posteriori. É também realizado, uma comparação entre o modelo misto proposto e o modelo exato introduzido por Gaba e VVinkler com o objetivo de verificar a eficiência da técnica aplicada. Uma ilustração com dados simulados é considerada
Title in English
Not available
Keywords in English
Not available
Abstract in English
Bayesian Inference for market research including answer errors is studied as a mbcture of two Bernoulli distributions. Since a formal Bayesian analysis leads to intractable calculations, a Markov Chain Monte Cano method with data augmentation is developed to obtain the summary and marginal posteriors. Were introduced a latent variables that indicates which component of the mixture gives the information with classification errors. Also, a Bayesian procedure based on the Beyes p-value and the total variation distance to measure the effect of the errors on the marginal posterior distribution. R is also, accomplished a comparison between the proposed mixed model and the exact model introduced by Gaba and Winkler with the objective of verifying the efficiency of the applied technique. An illustration with simulated data is considered.
 
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
2018-03-12
 
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.