BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Ta-Chu (Calvin) Kao  -- 
  CBL
DTSTART:20190319T160000Z
DTEND:20190319T170000Z
UID:TALK121774@talks.cam.ac.uk
CONTACT:Rodrigo Echeveste
DESCRIPTION:Ta-Chu (Calvin) Kao will be presenting:\n\n• Inferring singl
 e-trial neural population dynamics using sequential auto-encoders\n\n• C
 hethan Pandarinath\, Daniel J. O’Shea\, Jasmine Collins\, Rafal Jozefowi
 cz\, Sergey D. Stavisky\, Jonathan C. Kao\, Eric M. Trautmann\, Matthew T.
  Kaufman\, Stephen I. Ryu\, Leigh R. Hochberg\, Jaimie M. Henderson\, Kris
 hna V. Shenoy\, L. F. Abbott & David Sussillo\n\n• Nature Methods 2018\n
 \n• https://www.nature.com/articles/s41592-018-0109-9\n\nAbstract: Neuro
 science is experiencing a revolution in which simultaneous recording of th
 ousands of neurons is revealing population dynamics that are not apparent 
 from single-neuron responses. This structure is typically extracted from d
 ata averaged across many trials\, but deeper understanding requires studyi
 ng phenomena detected in single trials\, which is challenging due to incom
 plete sampling of the neural population\, trial-to-trial variability\, and
  fluctuations in action potential timing. We introduce latent factor analy
 sis via dynamical systems\, a deep learning method to infer latent dynamic
 s from single-trial neural spiking data. When applied to a variety of maca
 que and human motor cortical datasets\, latent factor analysis via dynamic
 al systems accurately predicts observed behavioral variables\, extracts pr
 ecise firing rate estimates of neural dynamics on single trials\, infers p
 erturbations to those dynamics that correlate with behavioral choices\, an
 d combines data from non-overlapping recording sessions spanning months to
  improve inference of underlying dynamics.
LOCATION:Cambridge University Engineering Department\, CBL\, BE4-38 (http:
 //learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR