Não foi possível enviar o arquivo. Será algum problema com as permissões?
Diferenças
Aqui você vê as diferenças entre duas revisões dessa página.
Ambos lados da revisão anterior Revisão anterior | Última revisão Ambos lados da revisão seguinte | ||
projetos:ehlers:par [2007/09/07 11:17] ehlers |
projetos:ehlers:par [2007/09/20 11:24] ehlers |
||
---|---|---|---|
Linha 1: | Linha 1: | ||
- | === Resumo === | + | === Abstract === |
In this project we study problems of inference and forecasting in autoregressive models with periodic correlation from a Bayesian perspective. Normality and unimodality assumptions are rarely verified in practice and the usual approach is to try Box-Cox transformations to obtain approximate normality and stabilize the periodic variance. More recently, mixture models were developed to take into account asymmetry and multimodality. | In this project we study problems of inference and forecasting in autoregressive models with periodic correlation from a Bayesian perspective. Normality and unimodality assumptions are rarely verified in practice and the usual approach is to try Box-Cox transformations to obtain approximate normality and stabilize the periodic variance. More recently, mixture models were developed to take into account asymmetry and multimodality. |