Banca de QUALIFICAÇÃO: EVERTON DA COSTA

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
DISCENTE: EVERTON DA COSTA
DATA : 25/03/2024
LOCAL: Meet (remoto)
TÍTULO:

Doubly-bounded Time Series Modeling: New Perspectives and Applications

PALAVRAS-CHAVES:

Resumo: We consider modeling and forecasting hydro-environmental time series subject to seasonal fluctuations and temporary regimes of an atypical nature, such as prolonged droughts. These atypical regimes have become more frequent as a result of climate change. We use the \beta-ARMA class of dynamic models, which was developed for use with doubly bounded data. We examine two important aspects of this class of models: the accuracy of hypothesis tests based on asymptotic approximations and the choice of link function. In particular, we show that
two commonly used tests can provide inaccurate inferences if the estimation of the null model is done on the basis of the maximum number of individual conditional log-likelihoods and we introduce a new model selection criterion that can be used with non-nested models and, in particular, for selecting the link function. Based on tests and the new criterion, we model the useful volumes of three Brazilian hydroelectric power plant water reservoirs. These time series exhibit seasonal fluctuations and contain atypical periods due to intense and prolonged droughts.

PÁGINAS: 1
RESUMO:

Beta distribution, beta-ARMA model, forecasting, hydro-environmental variables, hypothesis test, link function, time series

MEMBROS DA BANCA:
Presidente - 1279737 - FRANCISCO CRIBARI NETO
Externo à Instituição - FÁBIO MARIANO BAYER - UFSC
Interno - 1170969 - KLAUS LEITE PINTO VASCONCELLOS