SSA and Fourier Filters can not be used naked ... should be combined with NST-MTI Time series demonstration ..
Or it should be shifted by the lag indicator at the same N ratio
Apart from it, using naked causes big damages.
is there anyone who thinks it is possible
can someone prove to me that the program of neural networks makes money?
in vain
I will spend time and money
there is no analysis of SSA,
there is no Fourier analysis,
there is no cause-effect relationship,
no Bayesian methods
I am angry :):)
How to "defeat the playing indicators" in neural networks?
I decided to put this issue in a separate topic, because many have taken up the decision of this issue.
Somebody tell me, is it possible to "defeat" the redrawing turkeys when using them in neural networks