search

mobeets / gaborMotionPulses

fitting cell and behavior STRFs to gabor motion pulses, using ASD
2 stars 1 forks source link

opposite pairs better for decoding in terms of noise corrs #102

Closed mobeets closed 9 years ago

mobeets commented 9 years ago

screen shot 2015-07-28 at 1 10 18 pm

mobeets commented 9 years ago

very basic statement: the positive trend we see between RF correlation and noise correlation for positive rfCorr could just as well have continued for negative rfCorr. the fact that it doesn't means that, on the whole, pairs from opposing pools have noise-correlations that make them better suited than pairs from the same pool for decoding the stimulus direction.

mobeets commented 9 years ago

moving to trello since this isn't really an issue