Volume Independent Analytics for Discovery of Unseen Correlations via Tensors (VIADUCT)

Award Information
Agency: Department of Defense
Branch: Air Force
Contract: FA8750-18-C-0153
Agency Tracking Number: F171-048-0464
Amount: $739,580.00
Phase: Phase II
Program: SBIR
Solicitation Topic Code: AF171-048
Solicitation Number: 2017.1
Timeline
Solicitation Year: 2017
Award Year: 2018
Award Start Date (Proposal Award Date): 2018-09-27
Award End Date (Contract End Date): 2020-09-27
Small Business Information
632 Broadway, Suite 803, New York, NY, 10012
DUNS: 022423854
HUBZone Owned: N
Woman Owned: N
Socially and Economically Disadvantaged: N
Principal Investigator
 Dr. Thomas Henretty
 (212) 780-0527
 henretty@reservoir.com
Business Contact
 Ms. Melanie Peters
Phone: (212) 780-0527
Email: peters@reservoir.com
Research Institution
N/A
Abstract
This Phase II effort continues development of Volume Independent Analytics for Discovery of Unseen Correlations via Tensors (VIADUCT). VIADUCT leverages recent advances in tensor decompositions as a form of unsupervised discovery and the proliferation of open source Big Data tools to address the lack of advanced analytics available to Air Force Activity Based Intelligence (ABI) analysts. This effort extends VIADUCT with (1) support for more precision tensor decomposition methods and improved tensor construction tools, (2) improved performance and scalability to open Big Data platforms, tools, and data storage standards, and (3) automated post-processing support for ranking and identifying activities of interest. With these enhancements, VIADUCT will provide an automated machine learning capability for discovering previously unknown patterns-of-life and patterns-of-activity in billion- to trillion scale multi-INT datasets.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government