BEGIN:VCALENDAR PRODID:-//Microsoft Corporation//Outlook 16.0 MIMEDIR//EN VERSION:2.0 METHOD:PUBLISH X-MS-OLK-FORCEINSPECTOROPEN:TRUE BEGIN:VTIMEZONE TZID:W. Europe Standard Time BEGIN:STANDARD DTSTART:16011028T030000 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 END:STANDARD BEGIN:DAYLIGHT DTSTART:16010325T020000 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3 TZOFFSETFROM:+0100 TZOFFSETTO:+0200 END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT CLASS:PUBLIC CREATED:20210330T030822Z DESCRIPTION:The AICommunityOWL is a private\, independent network of AI ent husiasts.\nIt was founded in 2020 by employees of Fraunhofer IOSB-INA\, th e OWL\nUniversity of Applied Sciences (TH OWL)\, the Centrum Industrial IT \n(CIIT) and Phoenix Contact. Together\, they believe in digital progress\ nthrough the use of machine learning. They want to create sustainable\nsol utions for the challenges of the future: industry\, mobility\, smart\nbuil dings and smart cities - and above all\, for people!\nThe Machine Learning Reading Group (MLRG) of the AICommunityOWL has the\ngoal to get a better understanding of current trends in machine learning\non a technical level. The target audience are researchers and\npractitioners in the field of ma chine learning. We read and discuss\ncurrent papers with a high media impa ct or prominent positioning (at\nleast orals) of the leading conferences\, e.g. NeurIPS\, ICML\, ICLR\,\nAISTATS\, UAI\, COLT\, KDD\, AAAI\, CVPR\, ACL\, or IJCAI. Attendees are\nexpected to have read (or skimmed) the pape rs that are going to be\npresented so as not to be thrown off by the notat ion or problem\nstatement and to be able to participate in informed discus sions related\nto the paper. Suggestions for future papers are encouraged\ , as are\nvolunteer presenters.\nWe hold our first online meeting (after a yearlong hiatus) on Tuesday\,\nMai 4th\, at 16:00 under the following lin k:\nhttps://th-owl.webex.com/th-owl/j.php?MTID=mf53ff8ce643877e687995920b3 e5\n68b0 \n \nTitle: \nRecent Breakthroughs in Mastering Complex Video Gam es with Deep\nReinforcement Learning\nAbstract:\nStarCraft was long consid ered an unsolvable game\, using AI methods.\nThis has been proven wrong by DeepMind in 2019 when their reinforcement\nlearning agent achieved Grandm aster level in StarCraft II. We want to\ndiscuss some of the technical asp ects of AlphaStar and also take a brief\nlook at other challenging games t ackled using reinforcement learning\nmethods.\nLinks:\nhttps://www.seas.up enn.edu/~cis520/papers/RL_for_starcraft.pdf\n(AlphaStar)\nhttps://arxiv.or g/pdf/1912.06680.pdf (OpenAI Dota)\nhttps://arxiv.org/pdf/1901.10995.pdf ( GoExplore)\nSpeakers: Arthur Müller (Fraunhofer IOSB-INA) und Andreas Bes ginow (TH\nOWL)\n \nFor questions or suggestions of topics\, feel free to contact\nmarkus.lange-hegermann@th-owl.de\n \n \n DTEND;TZID="W. Europe Standard Time":20210504T180000 DTSTAMP:20210330T030822Z DTSTART;TZID="W. Europe Standard Time":20210504T160000 LAST-MODIFIED:20210330T030822Z LOCATION:https://th-owl.webex.com/th-owl/j.php?MTID=mf53ff8ce643877e6879959 20b3e568b0 PRIORITY:5 SEQUENCE:0 SUMMARY;LANGUAGE=en-us:Machine Learning Reading Group TRANSP:OPAQUE UID:040000008200E00074C5B7101A82E00800000000C032D3872225D701000000000000000 010000000F76A5A95D9C26742B112C977B34EFE16 X-ALT-DESC;FMTTYPE=text/html:

The AICommunityOWL is a private\, i ndependent network of AI enthusiasts. It was founded in 2020 by employees of Fraunhofer IOSB-INA\, the OWL University of A pplied Sciences (TH OWL)\, the Centrum Industrial IT (CIIT) and Phoenix Co ntact. Together\, they believe in digital progress through the use of mach ine learning. They want to create sustainable solutions for the challenges of the future: industry\, mobility\, smart buildings and smart cities - a nd above all\, for people!

The Machine Learning Reading Gro up (MLRG) of the AICommunityOWL has the goal to get a better understanding of current trends in machine learning on a tech nical level. The target audience are researchers and practitioners in the field of machine learning. We read and discuss current papers with a high media impact or prominent positioning (at least orals) of the leading conf erences\, e.g. NeurIPS\, ICML\, ICLR\, AISTATS\, UAI\, COLT\, KDD\, AAAI\, CVPR\, ACL\, or IJCAI. Attendees are expected t o have read (or skimmed) the papers that are going to be presented so as n ot to be thrown off by the notation or problem statement and to be able to participate in informed discussions related to the paper. Suggestions for future papers are encouraged\, as are volunteer presenters.

We hold our first online meeting (after a yearlong hiatus) on Tuesday\, Mai 4th\, at 16:00 under the following link: https://th-owl.webex.com/th-owl/j.php?MTID=mf53ff8ce6 43877e687995920b3e568b0

 \;

< p class=MsoNormal>Title:

Recent Breakthroughs in Mastering Complex Video Games with Deep Reinforcement Learning

Abstract:

StarC raft was long considered an unsolvable game\, using AI methods. This has b een proven wrong by DeepMind in 2019 when their reinforcement learning age nt achieved Grandmaster level in StarCraft II. We want to discuss some of the technical aspects of AlphaStar and also take a brief look at other challenging games tackled using reinforcement learn ing methods.

Links:

https://www.seas.upenn.edu/~cis520/papers/RL_for_starcraft.pdf (AlphaStar)

https: //arxiv.org/pdf/1912.06680.pdf (OpenAI Dota)

https://arxiv.org/pdf/ 1901.10995.pdf (GoExplore)

Sp eakers: Arthur Müller (Fraunhofer IOSB-INA) und Andreas Besginow (TH OWL)

 \;

For quest ions or suggestions of topics\, feel free to contact markus.lange-hegermann@th-owl.de

  \;

X-MICROSOFT-CDO-BUSYSTATUS:BUSY X-MICROSOFT-CDO-IMPORTANCE:1 X-MICROSOFT-DISALLOW-COUNTER:FALSE X-MS-OLK-AUTOFILLLOCATION:FALSE X-MS-OLK-CONFTYPE:0 BEGIN:VALARM TRIGGER:-PT15M ACTION:DISPLAY DESCRIPTION:Reminder END:VALARM END:VEVENT END:VCALENDAR