BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20200329T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20201025T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260406T024019Z
UID:1602853200@ist.ac.at
DTSTART:20201016T150000
DTEND:20201016T160000
DESCRIPTION:Speaker: Surya Ganguli\nhosted by Andrew Saxe\nAbstract: Remark
 able advances in experimental neuroscience now enable us to simultaneously
  observe the activity of many neurons\, thereby providing an opportunity t
 o understand how the moment by moment collective dynamics of the brain ins
 tantiates learning and cognition.  However\, efficiently extracting such 
 a conceptual understanding from large\, high dimensional neural datasets r
 equires concomitant advances in theoretically driven experimental design\,
  data analysis\, and neural circuit modeling.  We will discuss how the mo
 dern frameworks of high dimensional statistics and deep learning can aid u
 s in this process.  In particular we will discuss: (1) how unsupervised t
 ensor component analysis and time warping can extract unbiased and interpr
 etable descriptions of how rapid single trial circuit dynamics change slow
 ly over many trials to mediate learning\; (2) how to tradeoff very differe
 nt experimental resources\, like numbers of recorded neurons and trials to
  accurately discover the structure of collective dynamics and information 
 in the brain\, even without spike sorting\; (3) deep learning models that 
 accurately capture the retina’s response to natural scenes as well as it
 s internal structure and function\; (4) algorithmic approaches for simplif
 ying deep network models of perception\; (5) optimality approaches to expl
 ain cell-type diversity in the first steps of vision in the retina. Papers
 :A.H. Williams\, T.H. Kim\, F. Wang\, S. Vyas\, S.I. Ryu\, K.V. Shenoy\, M
 .J Schnitzer\, T.G. Kolda\, and S. Ganguli\, Unsupervised discovery of dem
 ixed\, low-dimensional neural dynamics across multiple timescales through 
 tensor components analysis\, Neuron 2018.J. Kadmon and S. Ganguli\, Statis
 tical mechanics of low-rank tensor decomposition\, Neural Information Proc
 essing Systems (NeurIPS) 2018.A. Williams\,\, B. Poole\, N. Maheswaranatha
 n\, A. Dhawale\, T. Fisher\, C. Wilson\, D. Brann\, E. Trautmann\, S. Ryu\
 , R. Shusterman\, D. Rinberg\, B. lveczky\, K. Shenoy\, S. Ganguli\, Disco
 vering precise temporal patterns in large-scale neural recordings through 
 robust and interpretable time warping\, Neuron\, 2019.O.I. Rumyantsev\, J.
 A. Lecoq\, J.C. Savall\, H. Zeng\, S. Ganguli\, M.J. Schnitzer\, Fundament
 al limits of information encoding by sensory cortical neural ensembles\, N
 ature 2020.E.M. Trautmann\, S. Lahiri\, S.D. Stavisky\, K.C. Ames\, M.T. K
 aufman\, S.I. Ryu\, S. Ganguli\, and K.V. Shenoy\, Accurate estimation of 
 neural population dynamics without spike sorting\, Neuron\, 2019.L. McInto
 sh\, N. Maheswaranathan\, S. Ganguli\, S. Baccus\, Deep learning models of
  the retinal response to natural scenes\, Neural Information Processing Sy
 stems (NIPS) 2016.L. McIntosh\, N. Maheswaranathan\, S. Ganguli\, S. Baccu
 s\, Deep learning models reveal internal structure and diverse computation
 s in the retina under natural scenes\, www.biorxiv.org/content/early/2018/
 06/08/340943 (http://www.biorxiv.org/content/early/2018/06/08/340943).H. T
 anaka\, A. Nayebi\, N. Maheswaranathan\, L.M. McIntosh\, S. Baccus\, S. Ga
 nguli\, From deep learning to mechanistic understanding in neuroscience: t
 he structure of retinal prediction\, NeurIPS 2019.S. Deny\, J. Lindsey\, S
 . Ganguli\, S. Ocko\, The emergence of multiple retinal cell types through
  efficient coding of natural movies\, Neural Information Processing System
 s (NeurIPS) 2018.  
LOCATION:Online Talk crowdcast\, ISTA
ORGANIZER:
SUMMARY:Surya Ganguli: Theoretical and computational approaches to neurosci
 ence with complex models in high dimensions across multiple timescales: fr
 om perception to motor control and learning
URL:https://talks-calendar.ista.ac.at/events/2879
END:VEVENT
END:VCALENDAR