[Comp-neuro] Kolmogorov complexity

james bower bower at uthscsa.edu
Wed Aug 20 20:43:51 CEST 2008


On Aug 20, 2008, at 1:24 PM, Carson Chow wrote:

> My suggestion is that we hold Bard's neural Turing test challenge  
> and try to reproduce your Purkinje cell model with something  
> simpler.  The test is simply that a set of unknown (to the  
> challenger) inputs will be fed to the challenger model and the  
> output has to be within some agreed upon precision of your model's  
> output.  Obviously, we would have to feed it an infinite amount of  
> data to really show the two models produced the same output but at  
> least we could see if it could work for a finite set.


First, there have been several published efforts to reduce the PC to a  
single point neuron.  Typically these efforts are focused more on  
generate 'realistic' patterns of output (spike firing, sometimes  
timing too), rather than providing a basis for better understanding  
how patterns of input produce output.  As I mentioned yesterday, in  
addition to serving as current conduits (which is how they are usually  
thought of in reduced form), dendrites also provide the basis for  
local (non-linear) interactions related to the particular spatial  
temporal patterns of local inputs.  By definition, I don't see how  
those can be collapsed into a single compartment model.  Adding some  
preprocessing component to such a model for inputs, might be as  
complex as the dendrite itself.

However, your challenge with respect to the Purkinje cell is  
interesting for another reason.  there is no experimental evidence, in  
vivo, under "natural stimulation conditions" that the 150,000  
excitatory parallel fiber inputs to Purkinje cells produce any direct  
influence on the spiking of the soma.

So, here is your challenge - build me a single compartment model 0r  
some other point neuron - give it 150,000 non zero synaptic inputs,  
and have those inputs have no direct effect on the timing of spike  
output.  Good luck.

:-)

Jim




==================================

Dr. James M. Bower Ph.D.

Professor of Computational Neuroscience

Research Imaging Center
University of Texas Health Science Center -
-  San Antonio
8403 Floyd Curl Drive
San Antonio Texas  78284-6240

Main Number:  210- 567-8100
Fax: 210 567-8152
Mobile:  210-382-0553

CONFIDENTIAL NOTICE:
The contents of this email and any attachments to it may be privileged  
or
contain privileged and confidential information. This information is  
only
for the viewing or use of the intended recipient. If you have received  
this
e-mail in error or are not the intended recipient, you are hereby  
notified
that any disclosure, copying, distribution or use of, or the taking of  
any
action in reliance upon, any of the information contained in this e- 
mail, or
any of the attachments to this e-mail, is strictly prohibited and that  
this
e-mail and all of the attachments to this e-mail, if any, must be
immediately returned to the sender or destroyed and, in either case,  
this
e-mail and all attachments to this e-mail must be immediately deleted  
from
your computer without making any copies hereof and any and all hard  
copies
made must be destroyed. If you have received this e-mail in error,  
please
notify the sender by e-mail immediately.










More information about the Comp-neuro mailing list