Found 5 results.




dynamic programming and optimal control (vol. i+ii)

Author(s): D.P. Bertsekas
Venue: Book
Year Published: 2017
Keywords: optimal control, dynamic programming
Expert Opinion: The optimal control formulation and the dynamic programming algorithm are the theoretical foundation of many approaches on learning for control and reinforcement learning (RL). In brief, many RL problems can be understood as optimal control, but without a-priori knowledge of a model. Thus, many algorithms and understanding in RL and robot learning build on optimal control. The series of books by Bertsekas provide an excellent introduction and reference into this field. While the first volume addresses primarily classical (model-based) optimal control, the second volume treats approximate dynamic programming, which includes addressing optimal control/dynamic programming problems with sampling-based methods. While these books do not directly target (machine) learning techniques, the underlying principles are key for addressing learning in robotics, and I thus consider these books as absolutely fundamental for this area. (Coincidentally, the new edition of Bertsekas' textbooks, which is announced for this year, will be called "Reinforcement Learning and Optimal Control‚".)

reinforcement learning and optimal control

Author(s): Dimitri P. Bertsekas
Venue: Book
Year Published: 2019
Keywords: reinforcement learning, optimal control, dynamic programming, neural networks
Expert Opinion: an accessible take on reinforcement learning that pairs well with the classic and influential book(s) on Dynamic Programming by Bertsekas

autonomous helicopter control using reinforcement learning policy search methods

Author(s): J. Andrew Bagnell, Jeff G. Schneider
Venue: IEEE International Conference on Robotics and Automation (ICRA)
Year Published: 2001
Keywords: learning from demonstration, reinforcement learning, dynamic programming
Expert Opinion: One of the first real demonstrations of RL on an actual robot performing a complex control problem.