Safe robot execution in model-based reinforcement learning

Conference object English OPEN
Martínez, David; Alenyà, Guillem; Torras, Carme;
(2015)
  • Publisher: Institute of Electrical and Electronics Engineers
  • Related identifiers: doi: 10.1109/IROS.2015.7354295, handle: 10261/133222
  • Subject: :Informàtica::Intel·ligència artificial [Àrees temàtiques de la UPC] | :Cybernetics::Artificial intelligence::Learning (artificial intelligence) [Classificació INSPEC] | manipulators | planning (artificial intelligence). | learning (artificial intelligence)

Task learning in robotics requires repeatedly executing the same actions in different states to learn the model of the task. However, in real-world domains, there are usually sequences of actions that, if executed, may produce unrecoverable errors (e.g. breaking an obje... View more
Share - Bookmark