optimal genetic options
5 years ago #114841
pardon me if that has been discussed. Does thing like optimal genetic options settings exist or is very much individual?
Which parameters are the key to optimal genetic search ?
Does strict initial population conditions give better “breading” and shorten run time to acceptable result?
How long does it normally take to develop to (relatively) robust strategy via genetic evolution so it can be passed on to further testing ? i.e. on i5-6600K machine?
When I test something very specific like custom indicator or some specific “situation” in the market (short time period) I use very small population size and more generations since I expect good results very quickly while when I experiment without any specific bias I usually use population size of several thousand of members
I think it does not affect final results that much. Trading hours setting, number of trades per day limitation, selected building blocks and exit rules are much more important criteria
If you have setting that cannot provide trading edge with any setting imaginable then the process can take forever while achieving nothing
Viewing 2 posts - 1 through 2 (of 2 total)