|
Activity Number:
|
608
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #304407 |
|
Title:
|
Change-Point Analysis for Stochastic Systems with Poisson Inputs
|
|
Author(s):
|
Stergios B. Fotopoulos*+
|
|
Companies:
|
Washington State University
|
|
Address:
|
Dept. Management & Operations, Pullman, WA, 99164-4736,
|
|
Keywords:
|
Likelihood ratio ; Maximum likelihood estimate ; Ladder epochs ; Renewal measure ; Random walk with negative drift
|
|
Abstract:
|
We embed the change-point problem into a stochastic system in which the input process is Poisson by allowing the underlying inter-arrival exponential rates to change subsequent to an unknown index of time. Adapting Hinkley's set up of the change-point problem, we derive explicit computable expressions for the asymptotic distribution of the maximum likelihood estimate of the unknown change-point. The analytical approach is based on an extensive application of the theory of random walks hitting the half-line. Numerical computations carried out illustrate the elegance as well as accuracy of the derived distribution theory.
|
