Machine Learning for Robots: a Comparison of Diierent Paradigms

For robots to be truly exible, they need to be able to learn to adapt to partially-known or dynamic environments, to teach themselves new tasks, and to compensate for sensor and eeector defects. The problem of robot learning has been an intensively studied research topic over the last decade. In this paper we critically examine four major formulations of the robot learning problem: inductive concept learning, explanation-based learning, reinforcement learning, and evolutionary learning. We describe some well-known examples of systems that t under each formulation, and discuss their strengths and limitations.