BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:iCalendar-Ruby
BEGIN:VEVENT
CATEGORIES:Academics,Academic Seminar,Virtual
DESCRIPTION:Title: Learning Nonlinear Operators Using Deep Neural Netwo
rks for Diverse Applications\n\nSpeaker: Dr. Lu Lu\, University of Penn
sylvania\n\nAbstract: It is widely known that neural networks (NNs) are
universal approximators of continuous functions. However\, a less known b
ut powerful result is that a NN can accurately approximate any nonlinear co
ntinuous operator. This universal approximation theorem of operators is su
ggestive of the structure and potential of deep neural networks (DNNs) in l
earning continuous operators or complex systems from streams of scattered d
ata. In this talk\, I will present the deep operator network (DeepONet) to
learn various explicit operators\, such as integrals and fractional Laplac
ians\, as well as implicit operators that represent deterministic and stoch
astic differential equations. More generally\, DeepONet can learn multisca
le operators spanning across many scales and trained by diverse sources of
data simultaneously. We will demonstrate the effectiveness of DeepONet to
multiphysics and multiscale problems.
DTEND:20211015T190000Z
DTSTAMP:20220927T232218Z
DTSTART:20211015T180000Z
LOCATION:
SEQUENCE:0
SUMMARY:Mathematics Colloquium - VIRTUAL
UID:tag:localist.com\,2008:EventInstance_37935949310123
URL:https://events.towson.edu/event/mathematics_colloquium_-_virtual_6406
END:VEVENT
END:VCALENDAR