By what means does WFM select its recommended combination?
5 replies
Threshold
7 years ago #115297
How does WFM select its recommended combination?
Patrick
7 years ago #138056
it choose the best option (period which brings biggest profit) if i understand it right.
Threshold
7 years ago #138058
it choose the best option (period which brings biggest profit) if i understand it right.
nope
Thats probably because of your your fitness, and I assume it chooses the best fitness option (based on the entire sample of data) but this isn’t explained anywhere.
CMKCMK
7 years ago #138063
My understanding is that it selects the center of the “Green Patch” (the green patch area is made up of #Runs and OS% points where they pass the selection criteria)
In that sense, the larger the area of the “green patch” (Top View on the 3D plot) the more stable the strategy
the grass is greener on the other side 😉
Threshold
7 years ago #138072
I don’t think that’s true either from lots of experience using my personal fitness score it seems to never be the middle.
This is kind of why I asked, I don’t think anybody knows how WFM works because there is no documentation explaining how it picks the recommended result.
I’m pretty sure it picks the result based on the best whole sample’s fitness score.
Thomas? Mark?
tomas262
7 years ago #138183
Hello,
I think what you mean is mentioned in the article Using Walk-Forward Matrix to find the optimal reoptimization period
https://strategyquant.com/articles/walk%20forward%20matrix
When we found for example a 3×3 group of single Walk-Forward optimizations,we prefer to select the one that is in the middle of it.
StrategyQuant does this automatically for us (if we configured it properly and the result is displayed just below the final result….
Viewing 5 replies - 1 through 5 (of 5 total)