BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Neural Processes - Andrew Foong\, Stratis Markou and Sebastian Obe
 r (University of Cambridge)
DTSTART:20201202T110000Z
DTEND:20201202T123000Z
UID:TALK153904@talks.cam.ac.uk
CONTACT:Elre Oldewage
DESCRIPTION:Neural Processes (NPs) are a recently proposed method for usin
 g meta-learning to train neural networks to predict stochastic processes. 
 They can be applied in regression tasks that require prediction with uncer
 tainty in the small-data regime and fast test-time inference. Furthermore\
 , the meta-learning framework allows NPs to learn intricate structure in t
 he stochastic process directly from the data\, allowing them to be applied
  to image data. In this reading group\, we will introduce the basic NP arc
 hitecture and go through some of the many kinds of NP that have been propo
 sed since 2018\, including the attentive NP and convolutional NP.\n\nSugge
 sted reading:\n\nThe talk will largely follow the content of this blog\, w
 hich also includes code/pre-trained models: \nhttps://yanndubs.github.io/N
 eural-Process-Family\n\nConditional Neural Processes: https://arxiv.org/ab
 s/1807.01613\n\nNeural Processes: https://arxiv.org/abs/1807.01622
LOCATION:https://eng-cam.zoom.us/j/86068703738?pwd=YnFleXFQOE1qR1h6Vmtwbno
 0LzFHdz09
END:VEVENT
END:VCALENDAR