Summary: weekly meeting 20160309 (Steve, me) - mobeets/nullSpaceControl GitHub Wiki

list of idiosyncratic behavior by day

• e.g. 20131205 tradeoff in speed for accuracy (ang=0,45)

20120601, targetAngle=225

• how is progress continuing to go up, but trial length saturating

handling incorrect trials for non-correct metrics

read: http://www.jneurosci.org/content/36/4/1056.full.pdf

NEXT STEPS:

• pick days with saturated behavior

  • re-run hypotheses on these days

• get IME working

  • run CCA on IME's estimate of Nul and Row

• total presentation for the ones I've chosen:

  • behavior
  • per target behavior
  • model performance overall
  • model performance per target
  • CCA