Dados do Trabalho

Título

A DYNAMIC MODEL FOR DOUBLE BOUNDED TIME SERIES WITH CHAOTIC DRIVEN CONDITIONAL AVERAGES

Resumo

In this work we introduce a dynamic class of model for time series taking values on the unit interval and following a beta distribution. The proposed model follows a generalized linear model approach where the mean specification includes covariates and also an extra additive term presenting chaotic behavior. The extra term is defined through the iteration of a chaotic transformation and can present a wide variety of behaviors including short and long range dependence, fixed points, periodic behavior, attracting and/or repelling points, etc.

Palavras-chave

chaotic processes; time series; beta regression