|
Activity Number:
|
29
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #306183 |
|
Title:
|
On Hinkley's Estimator: Inference about the Change-Point
|
|
Author(s):
|
Stergios B. Fotopoulos*+ and Venkata Jandhyala
|
|
Companies:
|
Washington State University and Washington State University
|
|
Address:
|
Department of Management and Operations, Pullman, WA, 99164,
|
|
Keywords:
|
maximum likelihood estimation ; multivariate Gaussian vector ; random walks ; Wiener-Hopf factorization ; ladder epochs
|
|
Abstract:
|
Maximum likelihood method is applied to estimate the change-point of a distribution function associated with a sequence of independent random elements. Fluctuation theory of random walks is applied to show exact expressions for the limiting distribution of the maximum likelihood estimator of a change-point. The derived expressions are computationally accessible in the sense that one may compute the exact distribution of the change-point through an algorithmic approach on the basis of the expressions derived. In showing this, a new formula for the ultimate maximum, the maximum of the sequence of partial maxima of a random walk is exhibited. As an example, the methodology is illustrated for estimating the change-point in the mean vector or/and variance-covariance matrix of the multivariate normal distribution.
|
