Times series
[STAT2170]
[22.5h + 7.5h] 5 credits
This course is taught in the 1st semester.
Teacher(s): Rainer von Sachs
Language: French
Level: Second
Aims
Main themes
Content and teaching methods
1. Modelling time series data: an introduction
2. Linear processes - simple parametric models (ARMA)
3. Estimation and prediction of ARMA models
4. Box-Jenkins analysis - (S)ARIMA models
5. Non-linear processes - heteroscedastic (G)ARCH models - applications to modelling financial data
Methods
Basic models of linear time series will be treated in the first part. The data analysis, i.e. estimation of the model parameters for forecasting, will be based predominantly on Box-Jenkins methods. In the second part of the course some elements of modelling financial data with the more recently developed ARCH and GARCH models will be given and included into the practical part of the course (done with the S-Plus software).
Other information (prerequisite, evaluation (assessment methods), course materials recommended readings, ...)
A general knowledge of basic statistical concepts (on the
level of a first introductory course in statistics) is necessary.
Evaluation
The examination will be oral. An applied data analysis project
has to be prepared on the computer.
Teaching material
Course notes, von Sachs, R. and S. Van Bellegem, Script.
References :
Brockwell, P., Davis, R. : Introduction to Time Series and Forecasting. 1996, Springer, New York
Brockwell, P., Davis, R. : Times Series : Theory and Methods. 1991, Springer, New York
Gourieroux, Ch. : Modèles ARCH et applications financières. 1992, Economica, Paris