BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//TUC//Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Europe/Athens
TZNAME:EEST
DTSTART:19700329T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0300
TZNAME:EET
DTSTART:19701025T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CREATED:20221018T115354Z
LAST-MODIFIED:20221018T115354Z
DTSTAMP:20260607T144109Z
UID:1780832469@tuc.gr
SUMMARY:Ομιλία του καθ. Νίκου Σιδηρόπουλου "
 Canonical identification of autoregr
 essive nonlinear systems with applic
 ation to smart health monitoring"
LOCATION:Λ - Κτίριο Επιστημών/ΗΜΜΥ, 145Π-58
DESCRIPTION:https://www.tuc.gr/el/to-polytechnei
 o/ilektronikes-ypiresies/imerologio/
 imerologio-ekdiloseon-1?tx_tucevents
 2_tuceventsdisplay%5Baction%5D=show&
 tx_tucevents2_tuceventsdisplay%5Bcon
 troller%5D=Event&tx_tucevents2_tucev
 entsdisplay%5Bevent%5D=5852&cHash=05
 d6ac6f22fdc77322afea4a87a440e3\nAbst
 ract\n Nonlinear dynamical system id
 entification is an important problem
  with numerous applications, from en
 gineering to scientific computing, h
 ealth and structural monitoring, and
  medicine and the life sciences. The
  linear case has been extensively st
 udied, but the identification of non
 linear dynamics is far more challeng
 ing. Recent work has focused on spar
 se multivariate polynomial dictionar
 y models, or recurrent neural networ
 k (RNN) models. A key issue with the
  former is how to pick the dictionar
 y when you have no (or limited) info
 rmation about the underlying dynamic
 s. RNNs are mostly black-box models 
 that provide limited engineering ins
 ight and generally require lots of t
 raining data to attain good performa
 nce. In recent work, we have conside
 red the problem of learning a contin
 uously differentiable nonlinear func
 tion via canonical tensor decomposit
 ion in the Fourier domain. In this t
 alk we consider the problem of ident
 ifying autoregressive nonlinear syst
 ems within this canonical decomposit
 ion framework. We develop an adaptiv
 e algorithm that is capable of prova
 bly identifying high-order nonlinear
  AR systems with limited storage and
  computational resources. The algori
 thm is carefully optimized to avoid 
 expensive tensor computations. We de
 monstrate its effectiveness in ident
 ifying the dynamics of photoplethysm
 ography (PPG) signals that are readi
 ly acquired by smart watches, and di
 scuss various smart health applicati
 ons that are enabled by the proposed
  methodology. \n About the speaker\n
  N. Sidiropoulos is the Louis T. Rad
 er Professor of Electrical and Compu
 ter Engineering at the University of
  Virginia. He earned his Ph.D. in El
 ectrical Engineering from the Univer
 sity of Maryland–College Park, in 19
 92. He has served on the faculty of 
 the University of Minnesota, and the
  Technical University of Crete, Gree
 ce. His research interests are in si
 gnal processing, communications, opt
 imization, tensor decomposition, and
  factor analysis, with applications 
 in machine learning and communicatio
 ns. He received the NSF/CAREER award
  in 1998, the IEEE Signal Processing
  Society (SPS) Best Paper Award in 2
 001, 2007, and 2011, and his student
 s received four IEEE SPS conference 
 best paper awards. Sidiropoulos has 
 authored a Google Classic Paper in S
 ignal Processing (on multicast beamf
 orming), and his tutorial on tensor 
 decomposition is ranked #1 in Google
  Scholar metrics for IEEE Transactio
 ns in Signal Processing (TSP), and t
 ops the charts of the most popular /
  most frequently accessed TSP papers
  in IEEExplore. He served as IEEE SP
 S Distinguished Lecturer (2008-2009)
 , Vice President of IEEE SPS (2017-2
 019), and as chair of the IEEE Fello
 w evaluation committee of SPS (2020-
 2021). He received the 2010 IEEE Sig
 nal Processing Society Meritorious S
 ervice Award, and the 2013 Distingui
 shed Alumni Award from the ECE Depar
 tment of the University of Maryland.
  He is a Fellow of IEEE (2009) and a
  Fellow of EURASIP (2014). He receiv
 ed the EURASIP Technical Achievement
  Award in 2022. More information at 
 http://www.ece.virginia.edu/~nds5j/ 
 and Google Scholar.\n
STATUS:CONFIRMED
ORGANIZER;RSVP=FALSE;CN=TUC;CUTYPE=TUC:mailto:webmaster@tuc.gr
DTSTART:20221021T170000
DTEND:20221021T180000
TRANSP:OPAQUE
CLASS:DEFAULT
END:VEVENT
END:VCALENDAR