Backpropagation as litmus test for mappings?
P C Knox
pck at castle.ed.ac.uk
Fri Feb 19 04:34:10 EST 1993
mitchm at casbah.acns.nwu.edu (Mitchell Maltenfort) writes:
> A grad student in my lab was looking at various algorithms to model
>optimal synergies of muscle activation, even though it's questionable whether
>the body is really optimal or just settles for "close enough." He and I both
>lean towards the latter, so he is trying to show that the nervous system is
>not using any consistent optimization scheme by making comparisons between
>real data and model predictions.
I can't answer the question I'm afraid, but I am puzzled by the last
sentence above. There is good evidence to show that the nervous system
does compare real data and model predictions in motor systems. Hence
experiments on the sensation of "heaviness" in limb muscles depend on
the notion that the desired state of contraction, the "model", signaled
by corollary discharges, is compared with the actual state, signaled by
muscle afferent (ie 1a) feedback.
More information about the Neur-sci