BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:www.tcs.tifr.res.in/event/1501
DTSTAMP:20241231T065603Z
SUMMARY:Learning with Structured Tensor Decompositions
DESCRIPTION:Speaker: Anand D. Sarwate (Rutgers University\, USA)\n\nAbstrac
 t: \nMany measurements or signals are multidimensional\, or tensor-valued.
  Vectorizing tensor data for statistical and machine learning tasks often 
 results in having to fit a very large number of parameters. Using tensor d
 ecompositions to model such data can give a flexible and useful modeling f
 ramework whose complexity can adapt to the amount of data available. This 
 talk will introduce classical decompositions (CP\, Tucker) as well as more
  recent ones (tensor train\, block tensor decomposition\, and low separati
 on rank) and show how they can be used to learn scalable representations f
 or tensor-valued data and make predictions from tensor-valued data. Time p
 ermitting\, we will describe applications of these ideas as part of neural
  networks and federated learning.Note: This talk does not assume the audie
 nce has prior familiarity with tensor algebra.\nShort Bio: Anand D. Sarwat
 e received his Ph.D. in electrical engineering from UC Berkeley. He is a c
 urrently an Associate Professor at Rutgers and was previously a Research A
 ssistant Professor at TTI-Chicago and a postdoc at the ITA Center at UCSD.
  His research interests include information theory\, machine learning\, si
 gnal processing\, optimization\, and privacy and security. Dr. Sarwate is 
 a Distinguished Lecturer of the IEEE Information Theory Society for 2024--
 2025 and is on the Board of Governors of the IEEE Information Theory Socie
 ty.\n
URL:https://www.tcs.tifr.res.in/web/events/1501
DTSTART;TZID=Asia/Kolkata:20250107T160000
DTEND;TZID=Asia/Kolkata:20250107T170000
LOCATION:HBA Foyer
END:VEVENT
END:VCALENDAR