Decision Making in Uncertainty: Modeling and Testing Multiple Theories
Abstract
In the proposed research, the performer will address the specific problem of how to facilitate operator decision-making when confronted with multiple possible courses of action, attempting to satisfy multiple objective functions. Situations with non-dominated alternatives are of particular interest, because no one of the possible course of action satisfies the set of object functions distinctly better than others; any alternate course of action that improves standing on one objective function degrades some other objective. Thus there is no mathematical algorithm that will generate prediction of a clearly superior outcome. There is inherent uncertainty encountered in these situations, manifesting both as the ambiguity about the current or future values of certain variables, and the complexity of the interrelationships among those variables. This uncertainty results in a large-scale multi-dimensional decision space, in which the number of possible options is so vast that it is overwhelming. A decision support tool that provides so many options leads to disuse, due to the ~tyranny of choice~ (Schwarz, 2004). The performer will develop a computational model of decision making that links our existing task network modeling capability with cognitive process modeling. They will build several scenarios and compare the output. The modeling environment will be used to help develop meta-theory about this class of decision making and to plan laboratory experiments to test and refine those theories.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 30, 2016
- Source ID
- N000141612863
Entities
People
- Dennis Folds
Organizations
- Georgia Tech Applied Research Corporation
- Office of Naval Research
- United States Navy