Dempster-Shafer Theory Applied to Admissible Regions
Author(s)
Worthy, Johnny L., III
Holzinger, Marcus J.
Advisor(s)
Editor(s)
Collections
Supplementary to:
Permanent Link
Abstract
The admissible region approach is often used a bootstrap method to initialize a
Bayesian state estimation scheme for too-short-arc measurements. However, there
are ambiguities in how prior probabilities are assigned for states in the admissible
region. Several approaches have proposed methods to assign prior probabilities,
however there are inconsistencies in how the prior probabilities can be manipulated.
The application of Dempster-Shafer evidential reasoning theory to the admissible
region problem can avoid these ambiguities by eliminating the need to
make any assumptions on the prior probabilities. Dempster-Shafer theory also enables
the testing of the validity of the assumptions used to construct the admissible
region. This paper introduces Dempster-Shafer theory and formulates the admissible
region in terms of plausibility and belief which reduce to traditional Bayesian
probability once there is sufficient information in the system.
Sponsor
Date
2017-02
Extent
Resource Type
Text
Resource Subtype
Paper
Rights Statement
Unless otherwise noted, all materials are protected under U.S. Copyright Law and all rights are reserved