In article <1992Apr8.235705.26412 at organpipe.uug.arizona.edu> bill at NSMA.AriZonA.EdU (Bill Skaggs) writes:
>In article <1992Apr8.151118.4780 at watmath.waterloo.edu>
>mwtilden at watmath.waterloo.edu (Mark W. Tilden) writes:
>>>>During a recent neural-networks lecture, it was pointed out that
>>comp-sci types only model one type of human neuron, where animal (human)
>>bodys contain no less than six different types.
>>
I'd be interested in hearing more about this. What criteria were used
to define these classes? What was the context of the talk?
>>I know I could look this up, but what is a net for? Could someone
>>please post the names of these six, and, if possible, a short discription
>>of their physiological locations and function (so far as is known).
>>> I doubt that you could look this up, because it's nonsense.
>Animal bodies contain literally hundreds of different types of
>neuron. Even if you only distinguish on the basis of the
>chemical neurotransmitter used, there are still at least several
>dozen different types. The functional consequences of these
>differences have hardly begun to be explored by neural network
>theorists.
>> -- Bill
Amen! If there only were six types we'd probably know everything by
now. :-)
Except for specail restricted purposes (retina physiolgy, cerebellum
physiology, etc.) the notion of a few types of neurons or standard
circuitry has no credibility among neuroscientists.
John Edstrom | edstrom @ elmer.hsc.ucalgary.ca
--
RM 2104, HSc Building, Div. Neuroscience
U. Calgary School of Medicine, 3330 Hospital Drive NW
Calgary, Alberta T2N 3Y4
(403) 220 4493 (wk)
