Lecturer: Prof. Dr. Markus Bibinger
- Time and place: This will be a virtual or hybrid seminar. We plan to have a block seminar instead of weekly meetings. The first organizational introduction is scheduled on Tuesday October 26 at 1:00 pm (13 Uhr) via zoom. If you want to participate but if you have a conflict with this appointment, please contact the lecturer.
- Beginn: October 26 2021, 1:00 pm
- Vst.-Nr. 08050860
- Language: The seminar will be in English or in German depending on the participants.
Depending on the prior knowledge of the participants, several topics in the field of statistics for stochastic processes are available. The idea is to have one block on statistics for time series, i.e. discrete-time stochastic processes. The second block is on statistics for continuous-time processes. Statistics for continuous-time processes is either based on an idealized model with observations in continuous time or on discrete-time observations of a continuous-time stochastic process. The discussed models are in particular applied in financial econometrics for financial data (but not limited to).
Possible topics include
- Estimation of stationary processes (time series)
- AR processes and Yule-Walker estimator
- Fractional Brownian motion and estimation of the Hurst exponent
- Parameter estimation for the Ornstein-Uhlenbeck process
- Nonparametric drift estimation for continuous-time diffusions
- Volatility estimation based on high-frequency data
Requirements: For the second block the lecture stochastic processes. For the first block at least one of the lectures Stochastik 2 or statistical analysis (or time series analysis in economics).
- Brockwell, Peter J. & Davis, Richard A. (2009), Time Series: Theory and Methods, Springer, ISBN: 978-1-4419-0320-4
- Brockwell, Peter J. & Davis, Richard A. (2016), Introduction to Time Series and Forecasting, Springer, ISBN: ISBN 3-319-29854-2
- Reiß, Markus (2015), Lecture notes on statistics of stochastic processes
(1) If you want to participate you should attend the organizational introduction and register via WueStudy.
(2) There is a WueCampus group for this seminar (link).