[dorkbotdc-blabber] Bayesian theory of the brain in New Scientist

Alberto Gaitán alberto.gaitan at gmail.com
Tue Jun 3 09:56:00 EDT 2008


Thanks, Philip.

I'd occasionally (at time = t) "pull out" a zorse.

That is, I'd model a branch, at t, into two parallel probability 
sequences, one resulting in a prior probabilities-based horse and one 
based on how the horse sequence (at t) might change based on an 
instantaneous environmental, random, or ontological factor: our "zebra." 
  Then I'd somehow (genetically or similarly modeling a feedback loop 
between them both) cross those two, perhaps following the resulting 
hybrid lineage for N generations to come up with a zorse.

Is that utterly naive?

Alberto




On 6/2/08 5:42 PM, Philip Kohn wrote:
> Alberto Gaitán wrote:
>> On our topic of modeling creativity, here's an interesting take on 
>> modeling the human brain as a Bayesian engine:
>>
>> http://reverendbayes.wordpress.com/2008/05/29/bayesian-theory-in-new-scientist/ 
>>
>>
>>
>>
>> ........................................................................
>> .......dorkbot dc: people doing strange things with electricity.........
>> ................... http://dorkbot.org/dorkbotdc .......................
>> ........................................................................
>>
> Friston and Hinton are both at the top of my neuroscientist heroes list.
> Neither is easy to understand without a strong background in math, 
> neural networks and neurobiology.
> Friston is particularly difficult.  I think he does it on purpose, but 
> you need a whole Friston vocabulary
> even to understand his text, no less his equations (which are always 
> plentiful).
> 
> What Bayesian does is allow you to give information about how probable 
> you think each outcome is.
> Doctors always talk about looking for horses, not zebras, which means 
> you look for the most likely
> possibility before considering the rare diagnosis that also matchs the 
> data (or may even match it better).
> Bayesian biases the result based on these "prior" probabilities.
> 
> As far as creativity, maybe you turn it on its head, and pull out the 
> zebras sometimes?
> 
> P.S., Congrats Alberto!!!
> 
> 
> 
> 
> 
> ........................................................................
> .......dorkbot dc: people doing strange things with electricity.........
> ................... http://dorkbot.org/dorkbotdc .......................
> ........................................................................
> 


More information about the dorkbotdc-blabber mailing list