Nathan Kent

This is the home page of Nate Kent. I am currently a PhD student at the University of Rochester and I am working in the Robotics and Artificial Intelligence Laboritory.

My current research is into the use of probabilistic graphical models to infer the correct Central Pattern Generator (CPG) parameters for underactuated robots. The current project focuses on using evolutionary algorithms to learn sets of parameters for various kinematic configurations of a hexapod robot and then using these parameters as a basis for inferring the appropriate CPG parameters for other, unseen configurations. Future work is to include inferring parameters across multiple kinds of robots (e.g., bipedal and quadruped), learning hierarchies of CPG behaviors, as well as utilizing reinforcement learning techniques.