|
Activity Number:
|
483
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
| Abstract - #307664 |
|
Title:
|
Parametric and Nonparametric Flowgraph Models with Application to Bayesian Multistate Systems Reliability
|
|
Author(s):
|
Aparna V. Huzurbazar*+ and Brian J. Williams and David Collins
|
|
Companies:
|
University of New Mexico and Los Alamos National Laboratory and University of New Mexico
|
|
Address:
|
MSC 03 2150 University of New Mexico, Albuquerque, NM, 87131-0001,
|
|
Keywords:
|
recurrent event ; time-to-event ; covariates
|
|
Abstract:
|
Flowgraph models have been used for a variety of problems in multistate systems involving time-to-event data. This talk will focus on some recent extensions of these models. The first concerns modeling recurrent events and including covariates in traditional parametric flowgraph models. Recurrent event modeling is increasingly important in the area of component and system reliability both in terms of repair and replacement. Although the strength of flowgraph models is that they model observable waiting times, allow for a variety of different distributions to be used within a stochastic network, and allow for all orders of loops, their weakness is that to date they do not allow for continuous time covariates. The second concerns extending these models to use nonparametric methods.
|
