Print Email Facebook Twitter Accelerating reinforcement learning on a robot by using subgoals in a hierarchical framework Title Accelerating reinforcement learning on a robot by using subgoals in a hierarchical framework Author Van Vliet, B. Caarls, W. Schuitema, E. Jonker, P.P. Faculty Mechanical, Maritime and Materials Engineering Department Biomechanical Engineering Date 2011-11-03 Abstract Reinforcement learning is a way to learn control tasks by trial and error. Even for simple motor control tasks, however, this can take a long time. We can speed up learning by using prior knowledge, but this is not always available, especially for an autonomous agent. One way to add limited prior knowledge is to use subgoals, defining points that the controller should aim for on the way to reaching the real goal. In this study, we use the MAXQ hierarchical framework to specify subgoals. This decreased the learning time by a factor two on a robot leg step-up task and we show that tests on a real robot give similar results. The worse end performance that is a result of the reduced solution space can be partially canceled out by hierarchical greedy execution. To our knowledge, this is the first time the MAXQ framework is applied to a real robot. To reference this document use: http://resolver.tudelft.nl/uuid:eaf340c0-5798-43ab-bc9d-e27c6dcd4e94 Source BNAIC 2011: 23rd Benelux Conference on Artificial Intelligence, Ghent, Belgium, 3-4 November 2011 Part of collection Institutional Repository Document type conference paper Rights (c) 2011 The Author(s) Files PDF vanVliet_2011.pdf 1.41 MB Close viewer /islandora/object/uuid:eaf340c0-5798-43ab-bc9d-e27c6dcd4e94/datastream/OBJ/view