Summary: weekly mtg 20160127 (Byron, Steve, Matt, me) - mobeets/nullSpaceControl GitHub Wiki

To do

Big picture

  1. orthogonalize latents using svd(L) from FA. Are our error metrics sensitive to this rotation/scaling? i.e., do relative errors on hypotheses change?
  2. volitional: find volitional control space (in place of Row(M2) from Blk1) using svd(L) from FA (see below). In the future, somehow regress out sensory-components of latents during Blk1.
  3. habitual: why does it fail for certain kinematics? are errors bigger for these? check into mean(abs(angError)) vs. mean(angError).
  4. find minimum mean error, cov using cross-validation

Quick

  1. look at all 4 datasets
  2. check covariance errors
  3. recreate DAP stick plot
  4. make comparison plot that shows all hypotheses as columns, null-space/kinematics as rows