Time Series Analysis (Winter Semester 2009/10)
- Lecturer: JProf. Dr. Claudia Kirch
- Classes: Lecture (1081), Problem class (1082)
- Weekly hours: 4+2
| Schedule | ||
|---|---|---|
| Lecture: | Monday 9:45-11:15 | HS 101 (10.50) |
| Wednesday 11:30-13:00 | 702 (10.50) | |
| Problem class: | Friday 11:30-13:00 | HS 101 (10.50) |
| Lecturers | ||
|---|---|---|
| Lecturer | JProf. Dr. Claudia Kirch | |
| Office hours: Nach Vereinbarung. | ||
| Room Kollegiengebäude Mathematik (20.30) | ||
| Email: claudia.kirch@kit.edu | Problem classes | Dipl.-Math. Franziska Häfner |
| Office hours: Mittwoch 14-15 Uhr (Bitte per E-Mail anmelden.) | ||
| Room Allianz-Gebäude (05.20) | ||
| Email: franziska.lindner@kit.edu | ||
Requirements
Fundamental knowledge of probability theory, e.g. 'Stochastik 2'. Some basic knowledge in statistics is helpful.
Content
The lecture gives an introduction to time series analysis. Topics that are dealt with:
- Stationary Time Series
- * Trend and seasonal components
- * Autocorrelation and spectral measure
- * ARMA-models
- * Wold decomposition
- Statistic in the time domain of stationary processes
- * Prediction
- * Estimators for mean and autocorrelation function
- * Estimation in ARMA models
- Statistic in the frequency domain of stationary processes
- * The periodogram
- * Nonparametric spectral density estimators
- Extensions and financial time series
- * ARIMA and SARIMA models
- * GARCH models
If time allows, we will also discuss further topics such as unit-root-tests or multivariate time series.
Material
The material concerning lecture and exercises can be found in the Studierendenportal.
References
- Brockwell, Davis: Time Series: Theory and Methods
- Kreiß, Neuhaus: Einführung in die Zeitreihenanalyse
- Franke, Härdle, Hafner: Statistics of financial markets (dt. Ausgabe: Einführung in die Statistik der Finanzmärkte)