{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,15]],"date-time":"2025-11-15T10:24:24Z","timestamp":1763202264057},"publisher-location":"California","reference-count":0,"publisher":"International Joint Conferences on Artificial Intelligence Organization","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2019,8]]},"abstract":"<jats:p>Artificial Intelligence has seen several breakthroughs in two-player \nperfect information game.\u00a0 Nevertheless, Doudizhu, a three-player \nimperfect information game, is still quite challenging.\u00a0 In this paper, \nwe present a Doudizhu AI by applying deep reinforcement learning from \ngames of self-play.\u00a0 The algorithm combines an asymmetric MCTS on nodes \nof information set of each player, a policy-value network that \napproximates the policy and value on each decision node, and inference \non unobserved hands of other players by given policy.\u00a0 Our results show \nthat self-play can significantly improve the performance of our agent in\n this multi-agent imperfect information game.\u00a0 Even starting with a weak\n AI, our agent can achieve human expert level after days of self-play \nand training.<\/jats:p>","DOI":"10.24963\/ijcai.2019\/176","type":"proceedings-article","created":{"date-parts":[[2019,7,28]],"date-time":"2019-07-28T07:46:05Z","timestamp":1564299965000},"page":"1265-1271","source":"Crossref","is-referenced-by-count":37,"title":["DeltaDou: Expert-level Doudizhu AI through Self-play"],"prefix":"10.24963","author":[{"given":"Qiqi","family":"Jiang","sequence":"first","affiliation":[{"name":"SweetCode Inc, Beijing"}]},{"given":"Kuangzheng","family":"Li","sequence":"additional","affiliation":[{"name":"SweetCode Inc, Beijing"}]},{"given":"Boyao","family":"Du","sequence":"additional","affiliation":[{"name":"SweetCode Inc, Beijing"}]},{"given":"Hao","family":"Chen","sequence":"additional","affiliation":[{"name":"SweetCode Inc, Beijing"}]},{"given":"Hai","family":"Fang","sequence":"additional","affiliation":[{"name":"SweetCode Inc, Beijing"}]}],"member":"10584","event":{"number":"28","sponsor":["International Joint Conferences on Artificial Intelligence Organization (IJCAI)"],"acronym":"IJCAI-2019","name":"Twenty-Eighth International Joint Conference on Artificial Intelligence {IJCAI-19}","start":{"date-parts":[[2019,8,10]]},"theme":"Artificial Intelligence","location":"Macao, China","end":{"date-parts":[[2019,8,16]]}},"container-title":["Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence"],"original-title":[],"deposited":{"date-parts":[[2019,7,28]],"date-time":"2019-07-28T07:47:21Z","timestamp":1564300041000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ijcai.org\/proceedings\/2019\/176"}},"subtitle":[],"proceedings-subject":"Artificial Intelligence Research Articles","short-title":[],"issued":{"date-parts":[[2019,8]]},"references-count":0,"URL":"https:\/\/doi.org\/10.24963\/ijcai.2019\/176","relation":{},"subject":[],"published":{"date-parts":[[2019,8]]}}}