Beefy Boxes and Bandwidth Generously Provided by pair Networks
go ahead... be a heretic
 
PerlMonks  

RE: How do our brains work?

by princepawn (Parson)
on Oct 18, 2000 at 00:02 UTC ( [id://37229]=note: print w/replies, xml ) Need Help??


in reply to How do our brains work?

Hmm, well I have a Master's Degree in something called Computational Neuroscience, an enterprise devoted to looking at the nervous system as a computational device.

Regarding your statements about associative memory, if the title of your post limited the discussion to the memory retrieval function of the brain, then your assertion would then be 1/2 OK... 1/2OK because there are two types of memory: declarative and procedural. Declarative memory is: "my name is bob" Procedural memory is how you remember how to throw darts --- the set of procedures to do something. What you (and all other posts) were discussing was declarative memory.

One neat thing about the brain is fault tolerance, meaning, that unlike a computer program, it can still manage to perform tasks in spite of input or self degradation. That is to say, a Perl program won't run at all once it is missing one semicolon and all searches for Purl will fail all associative search engines, but any human would be able to deal with the poor input and still make the relation after that night of drunken-ness.

Also, memory plays important part in other brain tasks. Take Neural Robotics, where robots perform various tasks (reaching and grasping, walking, etc.) based understandings of how mammals do it. Because the more times you do it, the better you get. But another thing that the brain is doing during this task is motor refinement via visual feedback. Meaning, you are looking at the act and based on what your brain "sees", the brain then refines the motor commands. And the motor commands are largely just increases or decreases in excitation to the muscles. This is brain controlled, but certainly has very little to do with association.

Finally, most of the best programs which attempt to simulate brain function have had to make serious re-considerations based on real physiological evidence from the brain. Both biological and artificial neural networks up until last year both used fixed numbers of inputs to the neurons and attempted to store memory in weights. The artificial weights were simple numbers and the biological weights were more complex and biochemically based. But in both cases, they assumed fixed connections between cells. But G.G. Turrigiano of UCSD, in Thinking globally, acting locally: AMPA receptor turnover and synaptic strength. Neuron. 1998 Nov;21(5):933-5 (you have to pay for online access to this paper --- no href provided) showed that no brain cell has the same synapses permanently. This means that even though the brain has relatively constant memory store, the memory can't be in the synapses because the receptors are constantly turning over.

Log In?
Username:
Password:

What's my password?
Create A New User
Domain Nodelet?
Node Status?
node history
Node Type: note [id://37229]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this?Last hourOther CB clients
Other Users?
Others surveying the Monastery: (5)
As of 2024-03-29 11:45 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    No recent polls found