eprintid: 10038400 rev_number: 26 eprint_status: archive userid: 608 dir: disk0/10/03/84/00 datestamp: 2018-01-04 11:29:05 lastmod: 2021-09-23 22:37:11 status_changed: 2018-01-04 11:29:05 type: proceedings_section metadata_visibility: show creators_name: Anthony, T creators_name: Tian, Z creators_name: Barber, D title: Thinking Fast and Slow with Deep Learning and Tree Search ispublished: pub divisions: UCL divisions: B04 divisions: C05 divisions: F48 note: This version is the version of record. For information on re-use, please refer to the publisher’s terms and conditions. abstract: Sequential decision making problems, such as structured prediction, robotic control, and game playing, require a combination of planning policies and generalisation of those plans. In this paper, we present Expert Iteration (ExIt), a novel reinforcement learning algorithm which decomposes the problem into separate planning and generalisation tasks. Planning new policies is performed by tree search, while a deep neural network generalises those plans. Subsequently, tree search is improved by using the neural network policy to guide search, increasing the strength of new plans. In contrast, standard deep Reinforcement Learning algorithms rely on a neural network not only to generalise plans, but to discover them too. We show that ExIt outperforms REINFORCE for training a neural network to play the board game Hex, and our final tree search agent, trained tabula rasa, defeats MoHex, the previous state-of-the-art Hex player. date: 2017-11-01 date_type: published publisher: NIPS Proceedings official_url: https://papers.nips.cc/paper/7120-thinking-fast-and-slow-with-deep-learning-and-tree-search oa_status: green full_text_type: pub language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 1511482 lyricists_name: Anthony, Thomas lyricists_name: Barber, David lyricists_id: TWANT76 lyricists_id: DBARB05 actors_name: Allington-Smith, Dominic actors_id: DAALL44 actors_role: owner full_text_status: public series: Neural Information Processing Systems publication: Advances in Neural Information Processing Systems volume: 30 place_of_pub: Long Beach, CA, USA event_title: Neural Information Processing Systems 2017 institution: Neural Information Processing Systems issn: 1049-5258 book_title: Advances in Neural Information Processing Systems 30 (NIPS 2017) pre-proceedings editors_name: Guyon, I editors_name: Luxburg, UV editors_name: Bengio, S editors_name: Wallach, H editors_name: Fergus, R editors_name: Vishwanathan, S editors_name: Garnett, R citation: Anthony, T; Tian, Z; Barber, D; (2017) Thinking Fast and Slow with Deep Learning and Tree Search. In: Guyon, I and Luxburg, UV and Bengio, S and Wallach, H and Fergus, R and Vishwanathan, S and Garnett, R, (eds.) Advances in Neural Information Processing Systems 30 (NIPS 2017) pre-proceedings. NIPS Proceedings: Long Beach, CA, USA. Green open access document_url: https://discovery-pp.ucl.ac.uk/id/eprint/10038400/1/Barber_7120-thinking-fast-and-slow-with-deep-learning-and-tree-search.pdf