BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//GetJobTools//Events//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:ece-seminar-series-stratis-ioannidis-2026@getjobtools.com
DTSTAMP:20260924T111844Z
DTSTART:20260930T190000Z
DTEND:20260930T200000Z
SUMMARY:ECE Seminar Series - Stratis Ioannidis
DESCRIPTION:Title: Learning Saliency-Preserving Latent Representations\n\
 n \n\nAbstract: We introduce an algorithm for learning salient feature re
 presentations through the explicit decomposition of salient and non-salie
 nt features into separate spaces. We show that our algorithm promotes lea
 rning low-dimensional\, task-relevant features. We prove that the expecte
 d prediction deviation under input perturbations is upper-bounded by the 
 dimension of the salient subspace and the so-called Hilbert-Schmidt Indep
 endence Criterion between inputs and representations. This establishes a 
 link between robustness and latent representation compression in terms of
  the dimensionality and information preserved. Empirical evaluations on i
 mage classification tasks show that models trained with our algorithm pri
 marily rely on salient input components\, as indicated by reduced sensiti
 vity to perturbations affecting non-salient features\, such as image back
 grounds.\n\n \n\nBio: Stratis Ioannidis is a professor in the Electrical 
 and Computer Engineering department of Northeastern University\, in Bosto
 n\, MA\, where he also holds a courtesy appointment with the Khoury Colle
 ge of Computer Sciences. He received his B.Sc. (2002) in Electrical and C
 omputer Engineering from the National Technical University of Athens\, Gr
 eece\, and his M.Sc. (2004) and Ph.D. (2009) in Computer Science from the
  University of Toronto\, Canada. Prior to joining Northeastern\, he was a
  research scientist at the Technicolor research centers in Paris\, France
 \, and Palo Alto\, CA\, as well as at Yahoo Labs in Sunnyvale\, CA. He is
  the recipient of an NSF CAREER Award\, a Google Faculty Research Award\,
  a Facebook Research Award\, a Søren Buus Outstanding Research Award\, a
  Martin W. Essigmann Outstanding Teaching Award\, and several best paper 
 awards. His research interests span machine learning\, distributed system
 s\, networking\, optimization\, and privacy.\n\nOrganizer: University of 
 Delaware events\n\nDetails and registration: https://events.udel.edu/even
 t/ece
LOCATION:Gore Hall\, 103\, United States
URL:https://events.udel.edu/event/ece-seminar-series-stratis-ioannidis
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR