On-line regression algorithms for learning mechanical models of robots: A survey

With the emergence of more challenging contexts for robotics, the mechanical design of robots is becoming more and more complex. Moreover, their missions often involve unforeseen physical interactions with the environment. To deal with these difficulties, endowing the controllers of the robots with the capability to learn a model of their kinematics and dynamics under changing circumstances is becoming mandatory. This emergent necessity has given rise to a significant amount of research in the Machine Learning community, generating algorithms that address more and more sophisticated on-line modeling questions. In this paper, we provide a survey of the corresponding literature with a focus on the methods rather than on the results. In particular, we provide a unified view of all recent algorithms that outlines their distinctive features and provides a framework for their combination. Finally, we give a prospective account of the evolution of the domain towards more challenging questions.

► Most up-to-date model learning methods combine linear models over local regions.
► LWPR is the most used, its main feature comes from NIPALS.
► XCSF brings the condition space/prediction space distinction.
► Several other systems bring additional features.
► The domain now evolves towards models of interaction with objects/users.