Would calculating a rolling average over the last n results and comparing that with the total average give you a useful metric? where n is much smaller than total runs. It may give you a clue about the trend. Maybe then if the delta between the two is small it's stable/stabilizing?
Just a thought...
In reply to Re: Instrumenting a genetic algorithm.
by RichardK
in thread Instrumenting a genetic algorithm.
by BrowserUk
For: | Use: | ||
& | & | ||
< | < | ||
> | > | ||
[ | [ | ||
] | ] |