RSS 2.0 Feed

» Welcome Guest Log In :: Register

Pages: (5) < [1] 2 3 4 5 >   
  Topic: Creating CSI with NS, H T T H H H T H T T H H H H T T T< Next Oldest | Next Newest >  

Posts: 3669
Joined: Mar. 2008

(Permalink) Posted: Oct. 08 2012,08:43   

Last night I did some calculator fiddling and discovered that the highest scoring pattern is runs of four. There may be a specific highest scoring pattern that includes a mixture, but I suspect the optimum is mostly fours with one five.

Math isn't my strongest subject, so it's likely I have screwed this up. But it appears there is a very small set of qualifying strings. Not quite as small as Mung's set of one string. But close.

The fitness function is similar to that used in the travelling salesman problem, and the GA has the ability to sum many variables into one score that has no knowledge of the target.

If it is to be considered a model of reality, then reality must allow fitness gradients. What the model demonstrates is that many dimensions of fitness can be addressed simultaneously.

”let’s not make a joke of ourselves.”

Pat Robertson

  128 replies since Oct. 06 2012,18:57 < Next Oldest | Next Newest >  

Pages: (5) < [1] 2 3 4 5 >   

Track this topic Email this topic Print this topic

[ Read the Board Rules ] | [Useful Links] | [Evolving Designs]