{"id":696090,"date":"2020-10-05T11:00:18","date_gmt":"2020-10-05T18:00:18","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-event&p=696090"},"modified":"2021-02-19T11:13:17","modified_gmt":"2021-02-19T19:13:17","slug":"reinforcement-learning-day-2021","status":"publish","type":"msr-event","link":"https:\/\/www.microsoft.com\/en-us\/research\/event\/reinforcement-learning-day-2021\/","title":{"rendered":"Reinforcement Learning Day 2021"},"content":{"rendered":"

This event has now concluded. On-demand content is available on the <\/strong>Videos tab<\/strong><\/a>.<\/p>\n

Previous events:
\nRL Day 2019<\/a>
\nRL Day 2018<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

This virtual reinforcement learning workshop will feature talks by a number of outstanding speakers whose research covers a broad swath of the topic, from statistics to neuroscience, from computer science to control.<\/p>\n","protected":false},"featured_media":705232,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"msr_startdate":"2021-01-14","msr_enddate":"2021-01-14","msr_location":"Virtual","msr_expirationdate":"","msr_event_recording_link":"","msr_event_link":"","msr_event_link_redirect":false,"msr_event_time":"","msr_hide_region":false,"msr_private_event":false,"footnotes":""},"research-area":[13556],"msr-region":[256048],"msr-event-type":[197944],"msr-video-type":[],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[],"msr-impact-theme":[],"class_list":["post-696090","msr-event","type-msr-event","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-region-global","msr-event-type-hosted-by-microsoft","msr-locale-en_us"],"msr_about":"This event has now concluded. On-demand content is available on the <\/strong>Videos tab<\/strong><\/a>.\r\n\r\nPrevious events:\r\nRL Day 2019<\/a>\r\nRL Day 2018<\/a>","tab-content":[{"id":0,"name":"About","content":"Reinforcement learning is the study of decision making with consequences over time. The topic draws together multi-disciplinary efforts from computer science, cognitive science, mathematics, economics, control theory, and neuroscience. The common thread through all of these studies is: how do natural and artificial systems learn to make decisions in complex environments based on external, and possibly delayed, feedback.\r\n\r\nThis virtual workshop featured talks by a number of outstanding speakers whose research covers a broad swath of the topic, from statistics to neuroscience, from computer science to control. A key objective was to bring together the research communities of all these areas to learn from each other and build on the latest knowledge.\r\n

<\/div>\r\n

Committee Chairs<\/h3>\r\nAkshay Krishnamurthy<\/a>, Microsoft Research\r\nChing-An Cheng<\/a>, Microsoft Research\r\nDipendra Misra<\/a>, Microsoft Research\r\nIda Momennejad<\/a>, Microsoft Research\r\nRobert Loftin<\/a>, Microsoft Research\r\n
<\/div>\r\n

Microsoft\u2019s Event Code of Conduct<\/h3>\r\nMicrosoft\u2019s mission is to empower every person and every organization on the planet to achieve more. This includes virtual events Microsoft hosts and participates in, where we seek to create a respectful, friendly, and inclusive experience for all participants. As such, we do not tolerate harassing or disrespectful behavior, messages, images, or interactions by any event participant, in any form, at any aspect of the program including business and social activities, regardless of location.\r\n\r\nWe do not tolerate any behavior that is degrading to any gender, race, sexual orientation or disability, or any behavior that would violate Microsoft\u2019s Anti-Harassment and Anti-Discrimination Policy, Equal Employment Opportunity Policy, or Standards of Business Conduct<\/a>. In short, the entire experience must meet our culture standards. We encourage everyone to assist in creating a welcoming and safe environment. Please report<\/a> any concerns, harassing behavior, or suspicious or disruptive activity. Microsoft reserves the right to ask attendees to leave at any time at its sole discretion.\r\n
<\/div>\r\n
[msr-button text=\"Report a concern\" url=\"https:\/\/app.convercent.com\/en-us\/Anonymous\/IssueIntake\/LandingPage\/65d3b907-0933-e611-8105-000d3ab03673\" new-window=\"true\" ]<\/div>"},{"id":1,"name":"Agenda","content":"

This event has now concluded.<\/h3>\r\n

Thursday,\u202fJanuary 14, 2021<\/h2>\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n
Time (EST) <\/strong><\/td>\r\n Session <\/strong><\/td>\r\n<\/td>\r\n Speaker <\/strong><\/td>\r\n<\/tr>\r\n
10:00 AM-10:15 AM<\/td>\r\nWelcome Remarks<\/td>\r\n\"Portrait<\/td>\r\nAkshay Krishnamurthy<\/a>, Microsoft Research<\/td>\r\n<\/tr>\r\n
10:15 AM-11:00 AM<\/td>\r\nNew Advances in Hierarchical Reinforcement Learning<\/td>\r\n\"Portrait<\/td>\r\nDoina Precup<\/a>, McGill University<\/td>\r\n<\/tr>\r\n
11:00 AM-11:45 AM<\/td>\r\nReinforcement Learning Debate: The State of RL and The Theory-Practice Divide<\/td>\r\n\"Portrait\r\n
<\/div>\r\n\"Portrait<\/a><\/td>\r\n
John Langford<\/a>, Microsoft Research\r\n
11:45 AM-12:15 PM<\/td>\r\nBreak<\/td>\r\n<\/td>\r\n<\/td>\r\n<\/tr>\r\n
12:15 PM-1:45 PM<\/td>\r\nVirtual Poster Presentations<\/td>\r\n<\/td>\r\n<\/td>\r\n<\/tr>\r\n
<\/td>\r\nInstance-Dependent Complexity of Contextual Bandits and Reinforcement Learning: A Disagreement-Based Perspective<\/td>\r\n<\/td>\r\nYunzong Xu, MIT<\/td>\r\n<\/tr>\r\n
<\/td>\r\nTaylor Expansion Policy Optimization<\/td>\r\n<\/td>\r\nYunhao Tang, Columbia University<\/td>\r\n<\/tr>\r\n
<\/td>\r\nProvably Efficient Policy Optimization with Thompson Sampling<\/td>\r\n<\/td>\r\nHaque Ishfaq, McGill University<\/td>\r\n<\/tr>\r\n
<\/td>\r\nActive Imitation Learning with Noisy Guidance<\/td>\r\n<\/td>\r\nKiant\u00e9 Brantley, University of Maryland<\/td>\r\n<\/tr>\r\n
<\/td>\r\nFinite-Time Analysis of Decentralized Stochastic Approximation with Applications in Multi-Agent and Multi-Task Learning<\/td>\r\n<\/td>\r\nSihan Zeng, Georgia Tech<\/td>\r\n<\/tr>\r\n
<\/td>\r\nMETA-Q-LEARNING<\/td>\r\n<\/td>\r\nRasool Fakoor, Amazon Web Services<\/td>\r\n<\/tr>\r\n
<\/td>\r\nToward the Fundamental Limits of Imitation Learning<\/td>\r\n<\/td>\r\nNived Rajaraman, UC Berkeley<\/td>\r\n<\/tr>\r\n
<\/td>\r\nMultitask Bandit Learning through Heterogeneous Feedback Aggregation<\/td>\r\n<\/td>\r\nZhi Wang, UC San Diego<\/td>\r\n<\/tr>\r\n
<\/td>\r\n\"It\u2019s Unwieldy and it Takes a Lot of Time.\u201d Challenges and Opportunities for Creating Agents in Commercial Games<\/td>\r\n<\/td>\r\nMikhail Jacob, Microsoft Research, Cambridge UK<\/td>\r\n<\/tr>\r\n
<\/td>\r\nA Framework for Robust Learning and Control of Nonlinear Systems with Large Uncertainty<\/td>\r\n<\/td>\r\nHoang Le, Microsoft Research, Redmond<\/td>\r\n<\/tr>\r\n
<\/td>\r\nLearning Dynamic Belief Graphs to Generalize on Text-Based Games<\/td>\r\n<\/td>\r\nEric Yuan, Microsoft Research, Montreal<\/td>\r\n<\/tr>\r\n
<\/td>\r\nFrugal Optimization for Cost-Related Hyperparameters<\/td>\r\n<\/td>\r\nQingyun Wu, Microsoft Research, NYC<\/td>\r\n<\/tr>\r\n
<\/td>\r\nImage Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels<\/td>\r\n<\/td>\r\nDenis Yarats, New York University<\/td>\r\n<\/tr>\r\n
<\/td>\r\nSelf Supervised Policy Adaptation During Deployment<\/td>\r\n<\/td>\r\nNicklas Hansen, Technical University of Denmark<\/td>\r\n<\/tr>\r\n
<\/td>\r\nMulti-Task Reinforcement Learning with Soft Modularization<\/td>\r\n<\/td>\r\nRuihan Yang, UC San Diego<\/td>\r\n<\/tr>\r\n
<\/td>\r\nContrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learning<\/td>\r\n<\/td>\r\nRishabh Agarwal, Google Research, and Mila Research<\/td>\r\n<\/tr>\r\n
<\/td>\r\nA Regret Minimization Approach to Iterative Learning Control<\/td>\r\n<\/td>\r\nKaran Singh, Princeton University<\/td>\r\n<\/tr>\r\n
<\/td>\r\nRMP2: A Differentiable Policy Class for Robotic Systems with Control-Theoretic Guarantees<\/td>\r\n<\/td>\r\nAnqi Li, University of Washington<\/td>\r\n<\/tr>\r\n
<\/td>\r\nGenerating Adversarial Disturbances for Controller Verification<\/td>\r\n<\/td>\r\nUdaya Ghai, Princeton University<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table>\r\n
<\/div>"},{"id":2,"name":"Call for papers","content":"

This event has now concluded.<\/h3>\r\n

Call for virtual poster session<\/h2>\r\nReinforcement learning as a field that studies the problem of sequential decision making with unknown and potentially long-term consequences. Reinforcement learning is a multi-disciplinary topic, bringing together diverse fields of study including computer science, cognitive science, mathematics, psychology, economics, control theory, and neuroscience. The common theme that connects these fields, and the core goal of reinforcement learning is the question: How do natural and artificial systems learn to make decisions in complex, unknown environments based on limited, noisy, and possibly delayed feedback?<\/em><\/strong>\r\n\r\nThis virtual workshop aims to bring together researchers from industry and academia to share and discuss recent advances, challenges, and future research directions for reinforcement learning. Our goal is to highlight emerging research opportunities for the reinforcement learning community, particularly those driven by the evolving need for robust decision making in practical applications. Reinforcement Learning Day 2021 will provide an opportunity for different research communities to learn from each other and build on the latest knowledge in reinforcement learning and related disciplines.\r\n

Invited speakers<\/h3>\r\nReinforcement Learning Day 2021 will feature invited talks and conversations with leaders in the field, including Yoshua Bengio<\/a> and John Langford<\/a>, whose research covers a broad array of topics related to reinforcement learning. For more details please see the agenda page.\r\n

Virtual poster session<\/h3>\r\nIn addition to our speaker program, Reinforcement Learning Day 2021 will include a virtual poster session, showcasing recent and ongoing research in all areas of reinforcement learning.\r\n\r\nWe invite you to submit posters on all topics related to reinforcement learning. Suggested topics include (but are certainly not limited to):\r\n