Re: [sl4] JAGI submission

From: Krekoski Ross (rosskrekoski@gmail.com)
Date: Tue Nov 25 2008 - 18:32:36 MST


Yes, and we are growing a population, not culling the entire
population save an individual every generation-- it may not be
immediately obvious which individual program to select. Also the
selection algorithm itself has a given complexity and needs to evolve
along with the system. For a trivial task this is of negligible
complexity, but for more complex tasks this could get quite
significant.

Ross

On Wed, Nov 26, 2008 at 9:13 AM, Nick Tarleton <nickptar@gmail.com> wrote:
> On Tue, Nov 25, 2008 at 8:12 PM, Nick Tarleton <nickptar@gmail.com> wrote:
>>
>> but picking out the maximal individual(s) according to some utility
>> function would only add a constant factor (complexity of the function).
>
> That is, add a constant factor to the O(log t) growth of population
> information.
> npt
>



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:03 MDT