|
Activity Number:
|
147
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
| Abstract - #305834 |
|
Title:
|
Estimating Failure Probability: Exploring Methods
|
|
Author(s):
|
Jason Overstreet*+ and Eric B. Howington and Alexia M. Athienitis-Makris
|
|
Companies:
|
Honeywell Inc. and Coastal Carolina University and Applied Health Outcomes
|
|
Address:
|
13350 U.S. Highway, Clearwater, FL, 33764,
|
|
Keywords:
|
reliability ; Weibull ; logistic regression ; failure probability ; censored
|
|
Abstract:
|
A common problem in the field of engineering and biology is estimating the probability of a "failure" based on an associated time variable. There are a number of methods available that allow the analyst to estimate the probability of failure with a prescribed level of certainty. Two popular techniques that are widely used to estimate failure probability based on a continuous variable (usually time) are Weibull analysis and binary logistic regression. This paper will present some unique results on the application of these and other methods that we hope will be of practical value to the practitioner.
|
