Scalable and Extensible Neural Sense-making Engine (SENSE)

ABSTRACT: In our phase-1 effort we proved feasible our Anti-Hebbian and Hebbian (AHaH) sparse-spiking framework for sense making applications. Demonstrated capabilities include optimal pattern classification, anomaly detection and complex temporal signal prediction. In the phase-2 effort we will develop a production ready Scalable and Extensible Neural Sense-making Engine (SENSE) built from clustered low-power single-board computers. We will furthermore develop the Network Adaptive Internet Intrusion Detection (NetAID) application that will utilize SENSE for its adaptive pattern recognition and anomaly detection capabilities. BENEFIT: The anticipated benefits of this effort include the development of a generic scalable sense-making system/utility capable of operating on arbitrary data streams. Generic capabilities include pattern recognition, prediction, anomaly detection, sensory fusion and motor control.