Na próxima sexta, dia 26 de abril pelas 11h00, vamos dar continuidade aos Seminários em Estatística e Ciência de Dados da Universidade Federal da Bahia. Nesta atividade teremos uma mistura entre seminários online e seminários presenciais. Este será Online (com transmissão no Auditório do IME/UFBA).
Temos o prazer de anunciar que a palestrante neste seminário é a Professora Nalini Ravishanker. A Nalini é Professora e Diretora do Programa de Graduação da Universidade de Connecticut, nos Estados Unidos. É também a Co-Editora Chefe do International Statistical Review e atual Presidente da International Society for Business and Industrial Statistics.
A participação neste seminário poderá ser presencial no Auditório do Instituto de Matemática e Estatística da UFBA ou via YouTube. Mais detalhes abaixo.
====================================================
Título: Hierarchical Dynamic Models for Multivariate Times Series of Counts
Resumo:
Discrete-valued time series modeling is emerging as an important area for many applications, as discussed in the recent CRC Handbook of Discrete-valued Time Series. Specifically, there is increasing interest in modeling univariate and multivariate time series of counts responses on several subjects as a function of subject-specific and/or time-dependent covariates. This talk presents a Bayesian framework for estimation and prediction by assuming a multivariate Poisson sampling distribution for the count responses and by fitting a hierarchical dynamic model which incorporates the temporal dependence as well as dependence between the components of the response vector. We illustrate this on ecology data to model count responses on different gastropod species. We also discuss a level correlated model (LCM) which enables us to account for association among the components of the response vector, possible overdispersion, and allows us combine different marginal count distributions and to build a hierarchical model for the vector time series of counts. We discuss the use of R-INLA for fast implementation of this flexible framework. We illustrate this on a marketing data set, by modeling the monthly prescription counts by physicians of a focal drug from a multinational pharmaceutical firm along with monthly counts of other competing drugs with sizable market share for the same therapeutic category.
Data: 11h00 do dia 26 de abril de 2019
Local: Auditório do Instituto de Matemática e Estatística da UFBA
Transmissão ao vivo online em: https://youtu.be/M9LJ2cyyTkw
