********************************************************************

                                                         Seminar
             Department of Systems Engineering and Engineering Management
                                  The Chinese University of Hong Kong

------------------------------------------------------------------------------------------

     
Title : Recent Developments in Long Memory Continuous-Time Series Analysis
     
Speaker : Dr. Tina Marquardt
    Munich University of Technology
     
Date : March 9th, 2007 (Friday)
     
Time : 4:30 p.m. - 5:30 p.m.
     
Venue : Room 513, MMW Engineering Building
    (Engineering Building Complex Phase 2)
    CUHK
     

------------------------------------------------------------------------------------------

Abstract:
 

The definition and properties of ARMA processes are reviewed. Since discrete-time series are often obtained by observing a continuous-time process at a discrete sequence of observation times, it is natural to model the underlying process as a continuous-time series. For this reason continuous-time ARMA (CARMA) models are introduced, which involve autocorrelation functions that show an exponential decrease over time. Hence, ARMA and CARMA processes belong to the family of short memory models. However, several measurements in hydrology, turbulence, finance, economics or telecommunications show long memory behaviour in the sense that they seem to require models, whose autocorrelation functions decay much less quickly. In this talk the fractionally integrated CARMA (FICARMA) models will be introduced, which belong to a class of long memory continuous-time stationary processes whose autocorrelation functions converge to zero at a power law. In particular, we develop a new approach to generate long memory moving aveage models by defining the class of fractional Levy processes which leads to a generalization of fractional Brownian motion. We also study the second order properties, sample path properties and consider integrals with respect to fractional Levy processes.


-------------------------------------------------------------------------------------------

Biography:
 

Tina Marquardt studied financial mathematics at the Munich University of Technology (TUM) in Germany. In 2004 she terminated her studies, writing her diploma thesis on "Continuous-time ARMA processes" with Peter Brockwell at the University of Melbourne (Australia) and the Colarado State University (USA). Back in Munich she started her Ph.D. studies under the supervision of Claudia Kluppelberg at the department of Mathematical Statistics (TUM). For her Ph.D. disseration in 2006 on "Fractional Levy processes, CARMA processes and related topics" she won several prizes. Currently she has a postdoc position at TUM working mainly in the fields of mathematical finance and (long memory) stochastic processes.


************************* ALL ARE WELCOME ************************

     
Host : Professor Xun Yu Zhou
Tel : (852) 2609-8320
Email : xyzhou@se.cuhk.edu.hk
     
Enquiries : Bolin Ding or Jeffrey Xu Yu
  : Department of Systems Engineering and Engineering Management
    CUHK
Website : http://www.se.cuhk.edu.hk/~seg5810
Email : seg5810@se.cuhk.edu.hk
     

********************************************************************