Charles River Analytics Inc., a developer of cutting-edge technologies for intelligent systems, announces a contract focused on probabilistic modeling. Charles River will develop the Probabilistic and Relational Inferences in Dynamic Environments (PRIDE) system for the Defense Advanced Research Projects Agency (DARPA).
“Under the PRIDE program, we are developing a framework for probabilistic modeling of complex situations,” said Dr. Avi Pfeffer, Principal Scientist at Charles River. Probabilistic modeling determines a situation’s unknown facts from known facts—a key reasoning task that is difficult because of the uncertainty and unpredictability of many factors, like sensors giving incomplete or noisy readings. Probabilistic modeling creates a model of a situation that captures all aspects of a situation and then uses probability theory to compute beliefs about the unknown facts from the known facts.
“Our approach has two components,” Dr. Pfeffer continued. “The first component is Probabilistic Relational Models (PRMs), an object-oriented representation that makes entities and relationships explicit. The second component is the Figaro probabilistic programming language, which supports a new kind of model-development process simplifying the challenging task of creating probabilistic models. Figaro—designed to represent many kinds of models, including PRMs—combines the power of functional programming with the ease of use of object-oriented programming.”
PRMs and Figaro are being used in a number of programs at Charles River, such as Assessment of Adversary Capability and Capacity via Intelligence Analysis (ACACIA).
Charles River has supported DARPA for over a decade in areas such as human computer interaction, situation assessment, predictive analysis, collaboration-support systems, and dynamic real-time operating systems. Current contracts with DARPA include:
- The Detection, Explanation, and Prediction of Emerging Network Developments (DEPEND) system as part of the GUARD DOG project
- Malware Analysis and Attribution using Genetic Information (MAAGI) as part of the Cyber Genome program
- Situation Awareness Modeling and Analysis Services for C2 Systems (SAMAS) as part of the RC2 program
This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No. W31P4Q-11-C-0083.
