Algorithmic Tools for Adversarial Games

Award Information
Agency:
Department of Defense
Amount:
$100,000.00
Program:
STTR
Contract:
FA9550-05-C-0110
Solitcitation Year:
2005
Solicitation Number:
N/A
Branch:
Air Force
Award Year:
2005
Phase:
Phase I
Agency Tracking Number:
F054-017-0278
Solicitation Topic Code:
AF05-T017
Small Business Information
SECURBORATION, INC.
695 Senderling Dr, Indialantic, FL, 32903
Hubzone Owned:
N
Woman Owned:
N
Socially and Economically Disadvantaged:
N
Duns:
038379579
Principal Investigator
 Lee Krause
 President
 (321) 591-9836
 lkrause@securboration.com
Business Contact
 Lynn Lehman
Title: VP Operation
Phone: (919) 244-3946
Email: llehman@securboration.com
Research Institution
 UNIV. OF CONNECTICUT
 Carol Welt
 438 Whitney Road Ext, Unit 113
Storrs, CT, 06269
 (860) 486-8704
 Nonprofit college or university
Abstract
Securboration, working with University of Connecticut researchers Dr. Eugene Santos Jr. is pleased to propose the Dynamic Adversarial Gaming Algorithm (DAGA). DAGA will focus on expanding adversarial gaming algorithms to support an agent based dynamic adversarial environment. DAGA will provide the following innovation to the adversarial gaming domain. 1) a natural mechanism to dynamically control the game based on current observables. 2) Support the interaction between agents through Web Ontology Language (OWL) based Common Operating Ontology. 3) Provided the ability for agent to split into multiple sub-agents as the population being represented diverges. 4) The use of Episodic learning to effect the group's behavior based on its experience over a period of time. Each of the listed innovation are required to support asymmetric adversarial games that represent the interaction between blue forces, red forces, and their interaction to influence green forces. The overriding goal of the DAGA service is to make accurate predication centered on a given actions ability to influence a "community of interest" to achieve a desired effect. The use of Bayesian Knowledge Fragments leverages the prediction strength of Bayesian base algorithms, along with the ability to account for prior knowledge in the prediction. In addition Bayesian Knowledge Fragments avoid the computationally cost and complexity of developing probability table typically associated with Bayesian approaches.

* information listed above is at the time of submission.

Agency Micro-sites

US Flag An Official Website of the United States Government