BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260601T162621EDT-12005MK7L0@132.216.98.100
DTSTAMP:20260601T202621Z
DESCRIPTION:Title: Neural network architectures for functional data analysi
 s\n\n \n\nAbstract:\n\nFunctional data is defined as any random variables 
 that assume values in an infinite precision domain\, such as time or space
 . In applications\, this data is usually discretely observed at some regul
 arly or irregularly-spaced points over the domain. In this talk\, we discu
 ss ways to adapt modern neural network architectures for the analysis of f
 unctional data. To do so\, we design new neural network layers in order to
  process functional data either as input\, output or both. First\, we prop
 ose the functional output layer\, which can be used to solve a multitude o
 f function-on-scalar regression problems in a non-linear way. The proposed
  layer provides a smooth representation of the output and we demonstrate h
 ow to regularize such a layer during the network training phase. Second\, 
 we propose a concept for functional weights that project functional data t
 o a scalar representation\, leading to a novel formulation for a functiona
 l input layer. We demonstrate how to combine both of these proposed functi
 onal layers to create a functional autoencoder. This model takes as input 
 the data in the form it is usually collected\, as discrete points over the
  domain\, and can be used for feature extraction and functional data smoot
 hing. We demonstrate the benefits of the proposed architectures with vario
 us experiments on simulated data and real data applications. We conclude w
 ith a brief discussion of ongoing work in the design of a functional convo
 lution layer that bridges the gap between the discrete convolution operati
 on and its continuous counterpart.\n\nSpeaker\n\nCédric Beaulac is a profe
 ssor of Statistics at the Université du Québec à Montréal. He completed a 
 postdoctoral fellowship at Simon Fraser University and at the University o
 f Victoria under the supervision of Farouk Nathou\, Jiguo Cao and Mirza Fa
 isal Beg and completed a Ph.D. in Statistical Sciences at the University o
 f Toronto under the supervision of Jeffrey S. Rosenthal. His research inte
 rests are at the intersection of machine learning and statistics. His rese
 arch focuses on the development of new models for image analysis and image
  generation by integrating machine learning models to functional data anal
 ysis.\n\nWebsite: https://cedricbeaulac.github.io\n\nhttps://mcgill.zoom.u
 s/j/89761165882\n\nMeeting ID: 897 6116 5882\n\nPasscode: None\n\n \n\n \n
 \n \n
DTSTART:20231020T193000Z
DTEND:20231020T203000Z
LOCATION:Room 1104\, Burnside Hall\, CA\, QC\, Montreal\, H3A 0B9\, 805 rue
  Sherbrooke Ouest
SUMMARY:Cédric Beaulac (Université du Québec à Montréal)
URL:/mathstat/channels/event/cedric-beaulac-universite
 -du-quebec-montreal-352099
END:VEVENT
END:VCALENDAR