BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:CIS Network Seminars: #ELLIS - Dr Luiz Chamon
DTSTART:20230202T140000
DTEND:20230202T150000
DTSTAMP:20260406T020624Z
UID:02825c0970c50f5505173b41a1f615169d9a6f88ce1751aec4d543bf
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr Luiz Chamon\nTitle: Learning under requirements\n\nAbstract
 : The transformative power of learning lies in automating the design of co
 mplex systems\, allowing us to go from data to operation with little to no
  human intervention. Today\, however\, learning does not incorporate requi
 rements organically\, which has led to data-driven solutions prone to tamp
 ering\, unsafe behavior\, and biased\, prejudiced actions. To realize its 
 autonomous engineering potential\, we must develop learning methods capabl
 e of satisfying statistical requirements beyond the training data. In this
  talk\, I will show when and how it is possible to learn under requirement
 s by developing the theoretical underpinnings of constrained learning. I w
 ill define constrained learning by extending the classical probably approx
 imately correct (PAC) framework and show that despite appearances\, constr
 ained learning is not harder than unconstrained learning. In fact\, they h
 ave essentially the same sample complexity. I will also derive a practical
  learning rule that under mild conditions can tackle constrained learning 
 tasks by solving only unconstrained empirical risk minimization (ERM) prob
 lems\, a duality that holds despite the lack of convexity. I will illustra
 te how these advances directly enable the data-driven design of trustworth
 y systems that adhere to fairness\, resource\, and safety specifications. 
 I will also discuss new advances in semi-infinite constrained learning and
  show how hybrid sampling/optimization algorithms can be used to tackle ch
 allenging problems in robust learning. These contributions suggest how we 
 can go beyond the current objective-centric learning paradigm towards a co
 nstraint-driven learning one.\n\nBio: Luiz F. O. Chamon received the B.Sc.
  and M.Sc. degrees in electrical engineering from the University of São P
 aulo\, São Paulo\, Brazil\, in 2011 and 2015 and the Ph.D. degree in elec
 trical and systems engineering from the University of Pennsylvania (Penn)\
 , Philadelphia\, in 2020. Until 2022\, he was a postdoctoral fellow at the
  Simons Institute of the University of California\, Berkeley. He is curren
 tly an independent research group leader at the University of Stuttgart\, 
 Germany\, and a faculty member of the International Max Planck Research Sc
 hool for Intelligent Systems (IMPRS-IS). In 2009\, he was an undergraduate
  exchange student of the Masters in Acoustics of the École Centrale de Ly
 on\, Lyon\, France\, and worked as an Assistant Instructor and Consultant 
 on nondestructive testing at INSACAST Formation Continue. From 2010 to 201
 4\, he worked as a Signal Processing and Statistics Consultant on a resear
 ch project with EMBRAER. He received both the best student paper and the b
 est paper awards at IEEE ICASSP 2020 and was recognized by the IEEE Signal
  Processing Society for his distinguished work for the editorial board of 
 the IEEE Transactions on Signal Processing in 2018. His research interests
  include optimization\, signal processing\, machine learning\, statistics\
 , and control.\n\nThe Center for Intelligent Systems at EPFL (CIS) is a co
 llaboration among IC\, ENAC\, SB\; SV and STI that brings together researc
 hers working on different aspects of Intelligent Systems.\n\nCollaboration
  and networking are crucial in the advancement of research and are key pil
 lars of the CIS’ activities. We are thrilled to announce a new series of
  talks: the “CIS Network Seminars”\, where we will invite experts from
  our network to share their work.\n\n \nThe CIS seminar will take place I
 n hybrid mode: Room INF 328 and by Zoom https://epfl.zoom.us/meeting/regis
 ter/u5Yrd-6qqjoiGd1PzhC_rzaq7Nd46MvohajV\n\n\nThursday February 2nd\, 2023
  from 2:00 to 3:00 pm\n\nNB: Video recordings of the seminars will be made
  available on our website and published on our social media pages
LOCATION:INF 328 https://plan.epfl.ch/?room==INF%20328 https://epfl.zoom.u
 s/meeting/register/u5Yrd-6qqjoiGd1PzhC_rzaq7Nd46MvohajV
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR