Workshop: Concluding Workshop
Time: Tue, December 02, 2008, 11:15-11:45
Speaker: Markus Hahn
We consider the problem of estimating the rate matrix governing a continuous-time finite-state Markov process given a number of (short) observation series. Such a situation arises naturally whenever there is some sort of break in the data; for financial data think of weekends for daily data or nights for tick-by-tick data. We propose to concatenate the observed series and to employ the emerging (non-Markov) process. We describe the bias arising if standard methods for Markov processes are used for the concatenated process, and provide a post-processing method to correct for this bias. This method works with any estimation approach (maximum likelihood / Bayesian / moment-based) and also applies to general models based on Markov jump processes where the underlying state process is not observed directly.
Presentation slides (pdf, 448 KB)
This page was made with 100% valid HTML & CSS - Send comments to Webmaster
Today's date and time is 11/28/22 - 15:16 CEST and this file (/specsem/sef/events/program/presentation.php) was last modified on 12/18/12 - 14:00 CEST