BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:PST8PDT BEGIN:DAYLIGHT DTSTART:20210314T030000 TZOFFSETFROM:-0800 TZOFFSETTO:-0700 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:PDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20201101T010000 TZOFFSETFROM:-0700 TZOFFSETTO:-0800 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:PST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20201129T181821Z UID:2E3C9763-40A7-41E1-BACA-FCFD2CBDFEB1 DTSTART;TZID=PST8PDT:20201126T160000 DTEND;TZID=PST8PDT:20201126T173000 DESCRIPTION:Please note new date of the event.\n\nDear All\,\n\nThe event i s postponed to November 26th\, 4pm due to organizational reasons. We apolo gize for any inconvenience. To those already registered the new link will be emailed. There is still time to register.\n\nThis free webinar is open to everyone. Please feel free to invite your friends and colleagues. We wo uld appreciate everyone to register using the link from this notice. The l ink to webinar will be shared to all registrants prior to the event.\n\nTh ank you\,\n\nIEEE Joint CS - RAS - SMC Chapter of Vancouver Section\n\n___ __________________________________________________________________________ ________________________________________\n\nTitle: Safety and robustness g uarantees with learning in the loop\n\nSpeaker: Prof. Dr. Nikolai Matni\, University of Pennsylvania\n\nAbstract:\nWe present recent progress toward s developing learning-based control strategies for the design of safe and robust autonomous systems. Our approach is to recognize that machine learn ing algorithms produce inherently uncertain estimates or predictions\, and that this uncertainty must be explicitly quantified (e.g.\, using non-asy mptotic guarantees of contemporary high-dimensional statistics) and accoun ted for (e.g.\, using robust control and optimization) when designing safe ty critical systems. In the first half of the talk\, we consider the optim al control of an unknown dynamical system\, and show that by integrating m odern tools from high-dimensional statistics and robust control\, we can p rovide end-to-end finite data robustness\, safety\, and performance guaran tees for learning and control. In the second half of the talk\, motivated by vision based control of autonomous vehicles\, we consider the complemen tary problem of controlling a known dynamical system for which partial sta te information\, such as vehicle position\, can only be extracted from hig h-dimensional data\, such as an image. Our approach is to learn a percepti on map from high-dimensional data to partial-state observation\, as well a s its corresponding error profile\, and then design a robust controller. W e show that jointly learning the perception map and error profile can be c ast as a robust regression problem\, and that this allows us to treat the perception map as a noisy position sensor\, which can then be integrated i nto a novel robust control synthesis procedure that has favorable safety a nd generalization properties. We conclude with our thoughts on future chal lenges and opportunities in the broad area of safe learning and control.\n \nBio:\nNikolai Matni is an Assistant Professor in the Department of Elect rical and Systems Engineering at the University of Pennsylvania\, where he is also a member of the Department of Computer and Information Sciences ( by courtesy)\, the GRASP Lab\, the PRECISE Center\, and the Applied Mathem atics and Computational Science graduate group. Prior to joining Penn\, Ni kolai was a postdoctoral scholar in EECS at UC Berkeley. He has also held a position as a postdoctoral scholar in the Computing and Mathematical Sci ences at Caltech. He received his Ph.D. in Control and Dynamical Systems f rom Caltech in June 2016. He also holds B.A.Sc. and M.A.Sc. in Electrical Engineering from the University of British Columbia\, Vancouver\, Canada. His research interests broadly encompass the use of learning\, optimizatio n\, and control in the design and analysis of safety-critical and data-dri ven cyber-physical systems. Nikolai was awarded the IEEE CDC 2013 Best Stu dent Paper Award (first ever sole author winner) and the IEEE ACC 2017 Bes t Student Paper Award (as co-advisor).\n\nSpeaker(s): Prof. Dr. Nikolai Ma tni \, \n\nAgenda: \nThursday November 26th\, 4pm-5:30pm\n\nWebinar. Free Event. Registration required.\n\nWebinar Link Info to be emailed upon regi stration.\n\nVancouver \, British Columbia\, Canada\, Virtual: https://eve nts.vtools.ieee.org/m/247891 LOCATION:Vancouver \, British Columbia\, Canada\, Virtual: https://events.v tools.ieee.org/m/247891 ORGANIZER:CSRASMCVancouver@gmail.com SEQUENCE:7 SUMMARY:Webinar (New Date): Safety and robustness guarantees with learning in the loop - Speaker: Prof. Dr. Nikolai Matni\, UPenn - Organizer: IEEE j oint Control\, Robotics\, and Cybernetics Chapter of the Vancouver Section URL;VALUE=URI:https://events.vtools.ieee.org/m/247891 X-ALT-DESC:Description: <br /><p>Please note new date of the event.</p>\n<p >&nbsp\;</p>\n<p>Dear All\,&nbsp\;</p>\n<p>The event is postponed to Novem ber 26th\, 4pm due to organizational reasons. We apologize for any inconve nience.&nbsp\;To those already registered the new link will be emailed. Th ere is still time to register.</p>\n<p>This free webinar is open to everyo ne. Please feel free to invite your friends and colleagues. We would appre ciate everyone to register using the link from this notice. The link to we binar will be shared to all registrants prior to the event.</p>\n<p>Thank you\,</p>\n<p>IEEE Joint CS - RAS - SMC Chapter of Vancouver Section</p>\n <p>_______________________________________________________________________ ______________________________________________</p>\n<p>Title: Safety and r obustness guarantees with learning in the loop</p>\n<p>Speaker: Prof. Dr. Nikolai Matni\, University of Pennsylvania</p>\n<p>&nbsp\;</p>\n<p>Abstrac t:<br />We present recent progress towards developing learning-based contr ol strategies for the design of safe and robust autonomous systems. Our ap proach is to recognize that machine learning algorithms produce inherently uncertain estimates or predictions\, and that this uncertainty must be ex plicitly quantified (e.g.\, using non-asymptotic guarantees of contemporar y high-dimensional statistics) and accounted for (e.g.\, using robust cont rol and optimization) when designing safety critical systems. In the first half of the talk\, we consider the optimal control of an unknown dynamica l system\, and show that by integrating modern tools from high-dimensional statistics and robust control\, we can provide end-to-end finite data rob ustness\, safety\, and performance guarantees for learning and control. In the second half of the talk\, motivated by vision based control of autono mous vehicles\, we consider the complementary problem of controlling a kno wn dynamical system for which partial state information\, such as vehicle position\, can only be extracted from high-dimensional data\, such as an i mage. Our approach is to learn a perception map from high-dimensional data to partial-state observation\, as well as its corresponding error profile \, and then design a robust controller. We show that jointly learning the perception map and error profile can be cast as a robust regression proble m\, and that this allows us to treat the perception map as a noisy positio n sensor\, which can then be integrated into a novel robust control synthe sis procedure that has favorable safety and generalization properties. We conclude with our thoughts on future challenges and opportunities in the b road area of safe learning and control.</p>\n<p>Bio:<br />Nikolai Matni is an Assistant Professor in the Department of Electrical and Systems Engine ering at the University of Pennsylvania\, where he is also a member of the Department of Computer and Information Sciences (by courtesy)\, the GRASP Lab\, the PRECISE Center\, and the Applied Mathematics and Computational Science graduate group. Prior to joining Penn\, Nikolai was a postdoctoral scholar in EECS at UC Berkeley. He has also held a position as a postdoct oral scholar in the Computing and Mathematical Sciences at Caltech. He rec eived his Ph.D. in Control and Dynamical Systems from Caltech in June 2016 . He also holds B.A.Sc. and M.A.Sc. in Electrical Engineering from the Uni versity of British Columbia\, Vancouver\, Canada. His research interests b roadly encompass the use of learning\, optimization\, and control in the d esign and analysis of safety-critical and data-driven cyber-physical syste ms. Nikolai was awarded the IEEE CDC 2013 Best Student Paper Award (first ever sole author winner) and the IEEE ACC 2017 Best Student Paper Award (a s co-advisor).</p>\n<p>&nbsp\;</p><br /><br />Agenda: <br /><p>Thursday No vember 26th\, 4pm-5:30pm</p>\n<p>Webinar. Free Event. Registration require d.</p>\n<p>Webinar Link Info to be emailed upon registration.&nbsp\;</p> END:VEVENT END:VCALENDAR