Intervalos de previsão bootstrap para modelos estruturais

This work is dedicated to the implementation of the methodology to calculate nonparametric bootstrap prediction intervals in state space form (SS), based on the work of Rodriguez and Ruiz (2009). The SS are an alternative way of rewriting the structural models, which decompose the time series in the...

Descripción completa

Detalles Bibliográficos
Autor: Thaize Vieira Martins
Tipo de recurso: tesis de maestría
Estado:Versión publicada
Fecha de publicación:2011
País:Brasil
Institución:Universidade Federal de Minas Gerais (UFMG)
Repositorio:Repositório Institucional da UFMG
Idioma:portugués
OAI Identifier:oai:repositorio.ufmg.br:1843/ICED-8HNHU8
Acceso en línea:http://hdl.handle.net/1843/ICED-8HNHU8
Access Level:acceso abierto
Palabra clave:sazonalidade
Forma de espaço de estados
filtro de Kalman
bootstrap
Bootstrap (Estatística)
Estatística
Analise multivariada
Series temporais
Análise espacial (Estatística)
Descripción
Sumario:This work is dedicated to the implementation of the methodology to calculate nonparametric bootstrap prediction intervals in state space form (SS), based on the work of Rodriguez and Ruiz (2009). The SS are an alternative way of rewriting the structural models, which decompose the time series in their non-observable components (level, trend and seasonality). Mainly, this work has the interest of extending the methodology proposed by Rodriguez and Ruiz (2009) to more complex structural models, implementing the algorithms in the Ox language and comparing the simulation results with the traditional method of constructing asymptotic prediction intervals and also with the parametric bootstrap procedure.