Aug 23, 2017 · 1 min read
Hello, thank you for your suggestion. Indeed the field of research is quite similar, but that area is cutting-edge research on “learning from demonstrations”, and it’s based on quite complex architectures, needing vaste computational time and efforts.
Here, we were exploring different techniques of intelligent control without any other information for the robot, to get a flavour of which architectures can rapidly and easily improve a robot performance with no supervision with respect to classic control theory.
