Sistemas complexos, séries temporais e previsibilidade

• Autor(a) principal: Carli, Henrique
• Data de Publicação: 2011
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UERJ
• Texto Completo: http://www.bdtd.uerj.br/handle/1/12891
• topic: Analysis of time series-Simulation methods; Differentiable dynamical systems; Probabilities; Chaotic behavior in systems; Análise de séries temporais - Métodos de simulação; Sistemas dinâmico diferenciais; Probabilidades; Comportamento caótico nos sistemas; CNPQ::CIENCIAS EXATAS E DA TERRA::MATEMATICA::GEOMETRIA E TOPOLOGIA::SISTEMAS DINAMICOS

For any observed system, physical or otherwise, one generally wishes to make predictions on its future evolution. Sometimes, very little is known about the system. If a time series is the only source of information on the system, prediction of the future values of the series requires a modelling of the system's (perhaps nonlinear) dynamical law. In particular, one is interested on the forecasting capabilities of the global approach to time series analysis. This can be a very complex and computationaly expensive procedure. In this work, we will concentrate in specific case: in some situations, the only information that we have about the system is a sequential data (or time serie). If we suppose that, for behind this data exists a dynamic with low dimensionality, we can use some techniques to reconstruct this dynamic. We are looking for build new techniques to improve the forecast for the known techniques over computional programming in Maple and C/C++.