tag:talks.cs.umd.edu,2005:/lists/31/feedRLSS2024-03-28T21:50:03-04:00tag:talks.cs.umd.edu,2005:Talk/27842021-02-24T11:24:30-05:002021-03-10T00:11:14-05:00https://talks.cs.umd.edu/talks/2784Autonomous Navigation of Atratospheric Balloons Using Reinforcement Learning and The History of Atari Games in Reinforcement Learning<a href="https://umd.zoom.us/j/2920984437">Marc Bellemare - Google Brain, MILA, McGill University</a><br><br>Wednesday, March 17, 2021, 4:00-5:15 pm<br><br><b>Abstract:</b> <p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);">Marc Bellemare created the Arcade Learning Environment (how RL interfaces with Atari games) and using it co-created deep reinforcement while at DeepMind. He'll be giving a talk on his recent Nature paper on controlling Loon balloons using RL, as well as the history of Atari games in reinforcement learning.</p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><a style="box-sizing: border-box; background-color: transparent; color: #cc0000; text-decoration-line: none; border-bottom: 1px solid transparent;" href="https://arxiv.org/abs/1207.4708">https://arxiv.org/abs/1207.4708</a></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><a style="box-sizing: border-box; background-color: transparent; color: #cc0000; text-decoration-line: none; border-bottom: 1px solid transparent;" href="https://www.nature.com/articles/nature14236">https://www.nature.com/articles/nature14236</a></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><a style="box-sizing: border-box; background-color: transparent; color: #cc0000; text-decoration-line: none; border-bottom: 1px solid transparent;" href="https://www.nature.com/articles/s41586-020-2939-8">https://www.nature.com/articles/s41586-020-2939-8</a></p><br>This talk is part of the following lists: <a href="https://talks.cs.umd.edu/lists/31">RLSS</a><br>tag:talks.cs.umd.edu,2005:Talk/27852021-02-24T11:26:06-05:002021-03-02T15:29:35-05:00https://talks.cs.umd.edu/talks/2785Deep Reinforcement Learning for Real-World Robotics<a href="https://umd.zoom.us/j/2920984437">Ilya Kuzovkin - Offworld.ai</a><br>Remote<br>Tuesday, March 2, 2021, 5:00-6:15 pm<br><br><b>Abstract:</b> <p><span style="font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);">OffWorld is developing a new generation of autonomous industrial robots to do the heavy lifting first on Earth, then on Moon, Mars and asteroids. We see reinforcement learning as one of major candidate technologies that could allow us to reach a high level of autonomy. While RL has achieved remarkable results in games and simulators, its adoption for real physical robots has been slow. In this talk we will go over a few projects we did at OffWorld that relate to applying RL on real robots, we then make the case that there is an apparent gap between RL community's aspirations to apply RL on real physical agents and its reluctance to move beyond simulators. To bridge this gap we introduce OffWorld Gym —a free access real physical environment and an open-source library that allows anyone to deploy their algorithms on a real robot using the familiar OpenAI gym ecosystem and without the burden of managing a real hardware system nor any knowledge of robotics.</span></p><br>This talk is part of the following lists: <a href="https://talks.cs.umd.edu/lists/31">RLSS</a><br>tag:talks.cs.umd.edu,2005:Talk/27862021-02-24T11:27:22-05:002021-03-30T21:20:03-04:00https://talks.cs.umd.edu/talks/2786Reinforcement learning for beating Super Smash Bros. Melee and Proving Mathematical Theorems<a href="https://umd.zoom.us/j/2920984437">Vlad Firoiu - DeepMind</a><br>Remote<br>Wednesday, March 31, 2021, 5:00-6:15 pm<br><br><b>Abstract:</b> <p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;">In the first half hour Vlad will discuss about hiswork on deep RL for Super Smash Bros. Melee: the road to building an AI that beats professional players, the challenges of making it a fair match between humans and machines, lessons learned along the way, and future directions smash bros. AI. In the second half of the talk Vlad will discuss his recent work on applying deep "RL" techniques to one of the most exciting application domains for AI: mathematics.</span></span></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;"> </span></span></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;">https://arxiv.org/abs/1702.06230</span></span></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;">https://arxiv.org/abs/1810.07286</span></span></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;">https://arxiv.org/abs/2103.03798</span></span></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"> </p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;"><span style="font-family: Source Sans Pro, Helvetica, sans-serif;"><span style="font-size: 16px;">https://umd.zoom.us/j/2920984437</span></span></span></span></p>
<p style="box-sizing: border-box; margin: 0px 0px 10px; font-family: 'Source Sans Pro', Helvetica, sans-serif; font-size: 16px; background-color: rgba(255, 255, 255, 0.8);"> </p><br>This talk is part of the following lists: <a href="https://talks.cs.umd.edu/lists/31">RLSS</a><br>tag:talks.cs.umd.edu,2005:Talk/27832021-02-24T11:13:17-05:002021-02-25T21:20:44-05:00https://talks.cs.umd.edu/talks/2783RIIT: Rethinking the Importance of Implementation Tricks in Multi-Agent Reinforcement LearningJian Hu & Seth Austin Harding - National Taiwan University, Taipei<br>Remote<br>Friday, February 26, 2021, 10:00-11:00 am<br><br><b>Abstract:</b> <p><span style="color: #222222; font-family: Arial, Helvetica, sans-serif; font-size: small;">In recent years, Multi-Agent Deep Reinforcement Learning (MADRL) has been successfully applied to various complex scenarios such as playing computer games and coordinating robot swarms. In this talk, we investigate the impact of “implementation tricks” for SOTA cooperative MADRL algorithms, such as QMIX, and provide some suggestions for tuning. In investigating implementation settings and how they affect fairness in MADRL experiments, we found some conclusions contrary to the previous work; we discuss how QMIX’s monotonicity condition is critical for cooperative tasks. Finally, we propose the new policy-based algorithm RIIT that achieves SOTA among policy-based algorithms.</span><br style="color: #222222; font-family: Arial, Helvetica, sans-serif; font-size: small;"><br style="color: #222222; font-family: Arial, Helvetica, sans-serif; font-size: small;"><a style="color: #1155cc; font-family: Arial, Helvetica, sans-serif; font-size: small;" href="https://arxiv.org/pdf/2102.03479.pdf" rel="noopener">https://arxiv.org/pdf/2102.<wbr></wbr>03479.pdf</a></p><br>This talk is part of the following lists: <a href="https://talks.cs.umd.edu/lists/2">CS Department</a> ⋅ <a href="https://talks.cs.umd.edu/lists/31">RLSS</a><br>tag:talks.cs.umd.edu,2005:Talk/31372022-03-29T08:39:16-04:002022-03-29T08:39:16-04:00https://talks.cs.umd.edu/talks/3137Generally Capable Reinforcement Learning AgentsJakob Bauer - DeepMind<br>Remote<br>Friday, April 1, 2022, 11:00 am-12:00 pm<br><br><b>Abstract:</b> <p><span style="color: #222222; font-family: Arial, Helvetica, sans-serif; font-size: small;">Artificial agents have achieved great success in individually challenging simulated environments, mastering the particular tasks they were trained for, with their behaviour even generalising to maps and opponents that were never encountered in training. In this talk we explore our recent work "Open-Ended Learning Leads to Generally Capable Agents" in which we create agents that can perform well beyond a single, individual task, that exhibit much wider generalisation of behaviour to a massive, rich space of challenges. We discuss the design of our environment spanning a vast set of tasks and how open-ended learning processes lead to agents that are generally capable across this space and beyond.</span><br style="color: #222222; font-family: Arial, Helvetica, sans-serif; font-size: small;"><br style="color: #222222; font-family: Arial, Helvetica, sans-serif; font-size: small;"><a style="color: #1155cc; font-family: Arial, Helvetica, sans-serif; font-size: small;" href="https://arxiv.org/abs/2107.12808" rel="noopener">https://arxiv.org/abs/2107.<wbr></wbr>12808</a></p><br>This talk is part of the following lists: <a href="https://talks.cs.umd.edu/lists/31">RLSS</a><br>