Anonymous user
Talk:Evolutionary algorithm: Difference between revisions
→How to make the algorithm scalable for larger target phrases?
Line 135:
:* I'll set <code>MUTATE</code> to 1000 and also try <code>COPIES</code> to 100 (currently it's 30).
:* Also, anyone please let me know if someone has already done this. It will save my poor laptop a few billion processing cycles. :) --[[User:Davidj|Davidj]] 12:45, 11 October 2011 (UTC)
:: i am running the pike version right now on a 5600 char string.after 300000 iterations it found 89% if only one character is changed per mutation. if it changes 10 characters per mutation then it only finds 48% at that point. so a low mutation rate is clearly more effective. as for scalability i am guessing the algorithm is linear: O(n).
::to speed it up some parallelization may be helpful. by mutating only one character at a time we know exactly which change caused a successful mutation. we thus can run multiple mutations at the same time and then combine the results.--[[User:EMBee|eMBee]] 13:45, 11 October 2011 (UTC)
| |||