{"id":705946,"date":"2020-11-16T00:14:01","date_gmt":"2020-11-16T08:14:01","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-group&#038;p=705946"},"modified":"2022-03-28T01:58:53","modified_gmt":"2022-03-28T08:58:53","slug":"deep-and-reinforcement-learning-group","status":"publish","type":"msr-group","link":"https:\/\/www.microsoft.com\/en-us\/research\/group\/deep-and-reinforcement-learning-group\/","title":{"rendered":"Deep and Reinforcement Learning Group"},"content":{"rendered":"<p>The Deep and Reinforcement Learning Group at Microsoft Research Asia pushes forward the research of deep learning and reinforcement learning from both algorithmic and practical aspects. Our interests include<\/p>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/deep-learning-and-representation-learning\/\">Deep representation learning<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, with focus on sequence to sequence learning and applications to <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/machine-translation-2\/\">machine translation<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/text-to-speech\/\">speech synthesis<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> and recognition, music understanding and composition, pre-training, etc.<\/li>\n<li>Deep structure learning, with focus on graph neural networks and applications to target discovery (in healthcare), protein modeling, drug design, etc.<\/li>\n<li>Deep reinforcement learning, with focus on distributional RL and offline RL and applications to logistics and supply chain management, etc.<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/learning-to-teach-and-automl\/\">Learning to teach and AutoML<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, transfer learning, generative models, and causal learning.<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<p>Our <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/dual-learning\/\">dual learning<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> and other techniques helped Microsoft achieve <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/blogs.microsoft.com\/ai\/chinese-to-english-translator-milestone\/\">human parity in Chinese-English<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> news translation in 2018, and win the <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/news.microsoft.com\/apac\/2019\/05\/22\/microsoft-research-asia-msra-leads-in-2019-wmt-international-machine-translation-competition\/\">first place for 8 translation tasks<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> in WMT 2019. We built the <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/news.microsoft.com\/apac\/features\/mastering-mahjong-with-ai-and-machine-learning\/\">world-best Mahjong AI<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, named <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/arxiv.org\/abs\/2003.13590\">Suphx<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, which achieved 10 DAN on the Tenhou platform in mid 2019. Our <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/arxiv.org\/abs\/1905.09263\">FastSpeech<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> model is the <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/techcommunity.microsoft.com\/t5\/azure-ai\/neural-text-to-speech-extends-support-to-15-more-languages-with\/ba-p\/1505911\">backbone of Azure TTS<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> and has supported 50+ languages\/locals and 80+ voices.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>The Deep and Reinforcement Learning Group at Microsoft Research Asia pushes forward the research of deep learning and reinforcement learning from both algorithmic and practical aspects. Our interests include Deep representation learning, with focus on sequence to sequence learning and applications to machine translation, speech synthesis and recognition, music understanding and composition, pre-training, etc. Deep [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_group_start":"","footnotes":""},"research-area":[13556],"msr-group-type":[243694],"msr-locale":[268875],"msr-impact-theme":[],"class_list":["post-705946","msr-group","type-msr-group","status-publish","hentry","msr-research-area-artificial-intelligence","msr-group-type-group","msr-locale-en_us"],"msr_group_start":"","msr_detailed_description":"","msr_further_details":"","msr_hero_images":[],"msr_research_lab":[199560],"related-researchers":[{"type":"user_nicename","display_name":"Tao Qin","user_id":33871,"people_section":"Section name 0","alias":"taoqin"},{"type":"user_nicename","display_name":"Li Zhao","user_id":36152,"people_section":"Section name 0","alias":"lizo"},{"type":"user_nicename","display_name":"Yingce Xia","user_id":37784,"people_section":"Section name 0","alias":"yinxia"},{"type":"user_nicename","display_name":"Shufang Xie","user_id":39886,"people_section":"Section name 0","alias":"shufxi"},{"type":"user_nicename","display_name":"Rui Wang","user_id":39880,"people_section":"Section name 0","alias":"ruiwa"},{"type":"user_nicename","display_name":"Jindong Wang","user_id":38455,"people_section":"Section name 0","alias":"jindwang"},{"type":"user_nicename","display_name":"Chang Liu","user_id":39889,"people_section":"Section name 0","alias":"changliu"},{"type":"user_nicename","display_name":"Renqian Luo","user_id":40441,"people_section":"Section name 0","alias":"renqianluo"},{"type":"user_nicename","display_name":"Guoqing Liu","user_id":40438,"people_section":"Section name 0","alias":"guoqingliu"},{"type":"user_nicename","display_name":"Junliang Guo","user_id":41188,"people_section":"Section name 0","alias":"junliangguo"}],"related-publications":[758137,758143,846901,853242,905004],"related-downloads":[],"related-videos":[],"related-projects":[708421,707674,558228,707419,707626,558237,558162,558135],"related-events":[],"related-opportunities":[],"related-posts":[],"tab-content":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/705946"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-group"}],"version-history":[{"count":5,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/705946\/revisions"}],"predecessor-version":[{"id":707434,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/705946\/revisions\/707434"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=705946"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=705946"},{"taxonomy":"msr-group-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group-type?post=705946"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=705946"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=705946"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}