The following list represents some of the clients who have either contracted for our services or sponsored research projects, primarily under the auspices of the Small Business Innovation Research (SBIR) program. Our research and development efforts have focused on applying neurocomputing, pattern recognition, and statistical modeling technologies to address specific applications. We acknowledge with deep gratitude the support of our agency sponsors in developing advanced technologies.

Martingale Research Clients / Sponsors Application
National Cancer Institute Robust classification research
UT Southwestern Medical Center Statistical Modeling - Randomized controlled clinical trials
U.S. Department of Veterans Affairs Statistical Modeling - Survey data analysis
National Institute on Alcohol Abuse and Alcoholism Database modeling and exploitation - statistical neural network
National Science Foundation Investigation of various soliton equations for use in stochastic filtering and control; Identification of the nonlinear Schrdinger equation as the ideal candidate. Detection and classification of structural fatigue crack growth using this wave-mechanical stochastic filter.
Defense Advanced Research Projects Agency Image understanding and machine cognition using a Bayesian neurocomputer architecture.
U.S. Army ARDEC - Picatinny Arsenal Track before detection of maneuvering targets. Precision beam stabilization of the ATB-1000 tank turret simulator in the presence of significant backlash. Feasibility of speech recognition using the Schrödinger equation in a stochastic filter.
U.S. Navy - Naval Surface Warfare Center Multi-target tracking and track-before-detect with the Parametric Avalanche stochastic filter, implemented on Transputers using Kernel Linda.
U.S. Air Force AFWAL: Wright-Patterson AFB Measured the mutual entropy of stimulus and response in H. Klopf's drive reinforcement learning algorithm using the Gibbs structure function.
National Aeronautics and Space Administration Nonlinear adaptive control of tethered satellites with sliding mode and Parametric Avalanche controllers. Achieved effective simulated control of in-plane oscillations with the PA controller.
U.S. Army CECOM - Communications and Electronics Command Studied the feasibility of combining a novel neural network with a generalized theory of modulation to achieve semi-automatic demodulation of covert signals.
U.S. Navy - Naval Research Laboratory Developed a general theory of modulation applied to non-cooperative communication systems. Using Hilbert space methods, found a unified description of angle, amplitude, and time-base modulation methods, in which unitary, Hermitian, and certain nonlinear operators correspond to messages.
Commercial Clients Interactive Video Database
On-line Help system
System Performance Modeling
Proposal Management
Other DoD Clients Neural Networks Research

Skip to navigation