Reducing Trials by Thinning-Out in Skill Discovery

作者: Hayato Kobayashi , Kohei Hatano , Akira Ishino , Ayumi Shinohara

DOI: 10.1007/978-3-540-75488-6_13

关键词:

摘要: In this paper, we propose a new concept, thinning-out, for reducing the number of trials in skill discovery. Thinning-out means to skip over such that are unlikely improve discovering results, same way as "pruning" search tree. We show our thinningout technique significantly reduces trials. addition, apply thinning-out discovery good physical motions by legged robots simulation environment. By using virtual can discover sophisticated is much different from initial motion reasonable amount

参考文章(17)
Min Sub Kim, William Uther, Automatic Gait Optimisation for Quadruped Robots ,(2003)
Damien Ernst, Arthur Louette, Introduction to Reinforcement Learning MIT Press. ,(1998)
Juan Cristóbal Zagal, Javier Ruiz-del-Solar, Learning to kick the ball using back to reality robot soccer world cup. pp. 335- 346 ,(2005) , 10.1007/978-3-540-32256-6_27
Hayato Kobayashi, Tsugutoyo Osaki, Eric Williams, Akira Ishino, Ayumi Shinohara, Autonomous Learning of Ball Trapping in the Four-Legged Robot League robot soccer world cup. pp. 86- 97 ,(2006) , 10.1007/978-3-540-74024-7_8
Tomoyuki Hiroyasu, Mitsunori Miki, Masaki Sano, Hisashi Shimosaka, Shigeyoshi Tsutsui, Jack Dongarra, Distributed Probabilistic Model-Building Genetic Algorithm Genetic and Evolutionary Computation — GECCO 2003. pp. 1015- 1028 ,(2003) , 10.1007/3-540-45105-6_112
Marco Zaratti, Marco Fratarcangeli, Luca Iocchi, A 3D Simulator of Multiple Legged Robots Based on USARSim robot soccer world cup. ,vol. 4434, pp. 13- 24 ,(2006) , 10.1007/978-3-540-74024-7_2
Peggy Fidelman, Peter Stone, The Chin Pinch: A Case Study in Skill Learning on a Legged Robot robot soccer world cup. pp. 59- 71 ,(2006) , 10.1007/978-3-540-74024-7_6