| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.91 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Sendo a prematuridade uma das principais causas associadas à mortalidade neonatal, em Portugal e no
Mundo, torna-se fundamental estudar a evolução da prevalência de nascimentos prematuros em
Portugal, num horizonte temporal de médio prazo (1989-2019).
Através da aplicação da Metodologia Box Jenkins foi possível escolher um modelo ARIMA adequado
à modelação dos nascimentos Prematuros, de Baixo Peso e de Muito Baixo Peso. Para tal, seguiram-se
as etapas propostas em estudos de análises de séries temporais: definição do problema, recolha de
informação, análise exploratória dos dados, escolha do melhor modelo, avaliação dos resultados e
previsão. A fase final de previsão com o modelo ARIMA consistiu num conjunto de previsões feitas
dentro e fora do horizonte temporal considerado. O poder preditivo do modelo estimado para cada uma
das séries foi, posteriormente, comparado com 2 modelos de previsão alternativos: o Modelo das Médias
Móveis e o Modelo de Alisamento Exponencial.
Prematurity is one of the main causes associated with neonatal mortality, not only in Portugal but globally, so it is essential to study the prevalence evolution of preterm births in Portugal in the medium term (1989-2019). Through the application of Box Jenkins Methodology, it was possible to choose an ARIMA model suitable to Preterm, Low-Weight and Extremely Low-Weight birth data. To make that attainable there were followed the steps proposed in time series studies: problem definition, gather information, data exploratory analysis, identify the best model, evaluation results and forecasting. The final step was applied within and outside the considered time series horizon. The predictive power of the estimated model for each time series was subsequently compared with 2 alternative forecasting models: the Moving Average Model and Exponential Smoothing Model.
Prematurity is one of the main causes associated with neonatal mortality, not only in Portugal but globally, so it is essential to study the prevalence evolution of preterm births in Portugal in the medium term (1989-2019). Through the application of Box Jenkins Methodology, it was possible to choose an ARIMA model suitable to Preterm, Low-Weight and Extremely Low-Weight birth data. To make that attainable there were followed the steps proposed in time series studies: problem definition, gather information, data exploratory analysis, identify the best model, evaluation results and forecasting. The final step was applied within and outside the considered time series horizon. The predictive power of the estimated model for each time series was subsequently compared with 2 alternative forecasting models: the Moving Average Model and Exponential Smoothing Model.
Descrição
Trabalho de Projeto de Mestrado, Matemática Aplicada à Economia e Gestão, 2022, Universidade de Lisboa, Faculdade de Ciências
Palavras-chave
Séries Temporais Nascimentos Prematuros ARIMA Métodos de Previsão Teses de mestrado - 2023
