Time Series (MP-CA)

Área

AC Matemática > UC Mestrados

Activa nos planos curriculares

Actuarial Science > Actuarial Science > 2º Ciclo > Unidades Curriculares Obrigatórias > Time Series

8ª Edição > 8ª Edição > 2º Ciclo > Unidades Curriculares Obrigatórias > Time Series

Nível

2º Ciclo (M)

Tipo

Não Estruturante

Regime

Semestral

Carga Horária

Aula Teórica (T): 0.0 h/semana

Aula TeoricoPrática (TP): 3.0 h/semana

Trabalho Autónomo: 129.0 h/semestre

Créditos ECTS: 6.0

Objectivos

On completion of this course, the student should be able to:
- Recognize and understand the main econometric models used in the analysis of time series.
- Understand the theoretical reasoning which led to the development of the most important univariate and multivariate models.
- Be familiar with the use of econometric software to carry out time series analysis.
- Develop critical thinking about empirical work with time series data.
- Be able to develop a forecasting study of different sets of variables and formulate statistical hypotheses of interest. Understand the limitations of the econometric methodology applied in the study

Programa

- Introduction to time series analysis. Fundamental concepts
- Models for stationary time series. Autoregressive Moving Average (ARMA) models
- Box-Jenkins methodology: model identification, estimation and diagnostic checking
- Models for nonstationary time series. Autoregressive Integrated Moving Average (ARIMA) models and unit root testing
- Forecasting using ARIMA models
- Seasonality and Seasonal ARIMA (SARIMA) models
- Conditional Heteroskedasticity time series models. ARCH/GARCH models
- Forecasting with exponential smoothing methods
- Multivariate Time Series Models

Metodologia de avaliação

Lectures will be theoretical and practical, starting on main empirical patterns found in time series as a basis to present statistical methods and models used to represent it. Core mathematical models for time series will be presented in a constructive way, but practical relevance of different models in terms of time series behavioural patterns and on the nature of implied forecast functions will also be strengthened. Using available software, models and modelling strategies will be applied on real time series data with emphasis in critical analysis as a function of purposes.
Students will be assessed based on a final exam (60%) and a practical computational test (40%) using R

Bibliografia

Principal