Seminar CogniGron - Dr. M. Cook

Title: Relational Processing in Neuromorphic Computing

Abstract:
Relational processing consists of using known relationships to make inferences from partial information. Well-known examples include satisfiability problems and graphical models. In our group, we design relational processing systems that are neuromorphic at a high level, following basic architectural principles found in the brain, and through collaborations with hardware designers we map our algorithms onto neuromorphic hardware. I will present some examples of neuromorphic relational processing that we have created to deal with sensory input, internal processing, and motor output.