BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Guillaume Hennequin (CBL
 )
DTSTART:20180717T150000Z
DTEND:20180717T160000Z
UID:TALK108337@talks.cam.ac.uk
CONTACT:Rodrigo Echeveste
DESCRIPTION:Guillaume Hennequin will cover:\n\n• Unsupervised Discovery 
 of Demixed\, Low-Dimensional Neural Dynamics across Multiple Timescales th
 rough Tensor Component Analysis\n\n• Alex H. Williams\, Tony Hyun Kim\, 
 Forea Wang\, Saurabh Vyas\, Stephen I. Ryu\, Krishna V. Shenoy\, Mark Schn
 itzer\, Tamara G. Kolda\, Surya Ganguli\n\n• Neuron (2018)\n\n• https:
 //doi.org/10.1016/j.neuron.2018.05.015\n\n\nAbstract: Perceptions\, though
 ts\, and actions unfold over millisecond timescales\, while learned behavi
 ors can require many days to mature. While recent experimental advances en
 able large-scale and long-term neural recordings with high temporal fideli
 ty\, it remains a formidable challenge to extract unbiased and interpretab
 le descriptions of how rapid single-trial circuit dynamics change slowly o
 ver many trials to mediate learning. We demonstrate a simple tensor compon
 ent analysis (TCA) can meet this challenge by extracting three interconnec
 ted\, low-dimensional descriptions of neural data: neuron factors\, reflec
 ting cell assemblies\; temporal factors\, reflecting rapid circuit dynamic
 s mediating perceptions\, thoughts\, and actions within each trial\; and t
 rial factors\, describing both long-term learning and trial-to-trial chang
 es in cognitive state. We demonstrate the broad applicability of TCA by re
 vealing insights into diverse datasets derived from artificial neural netw
 orks\, large-scale calcium imaging of rodent prefrontal cortex during maze
  navigation\, and multielectrode recordings of macaque motor cortex during
  brain machine interface learning.
LOCATION:Cambridge University Engineering Department\, CBL\, BE4-38 (http:
 //learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR