You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Thank you for the flattering assessment.
You're a master at evading answers.
So what do you want to suggest?
All right, well.
To prove something, you have to show something.
I'll open a non-subscription signal on Monday. Then we will talk.
Just compare.
No one's saying you're wrong.
Change the inputs to your system and see the results.
By the way, I'm interested in it too.
the reality is you don't know what the noise is for your TS
so by changing the filters, you're just changing the TS
I suggest you to close this useless topic with digital signal processing, which is not suitable for financial markets from the word "at all".
the reality is that you don't know what the noise is for your TS
so, changing the filters, you just change the TS
I propose to close this useless topic with digital signal processing, which is not suitable for financial markets from the word "at all"
does that sound convincing?
admin's words:
https://www.mql5.com/ru/forum/101311/page2#comment_2962042
Will that sound convincing?
admin's words:
https://www.mql5.com/ru/forum/101311/page2#comment_2962042
no. But this sounds convincing.
https://www.mql5.com/ru/forum/101311/page8#comment_2962102No. Now that sounds convincing.
https://www.mql5.com/ru/forum/101311/page8#comment_2962102OK
The problem with the Fourier series decomposition is that the problem is solved head-on, i.e. with noise
it must first be filtered out and only then decomposed
then apply a coefficient to each frequency, or a predictor, as is customary here
and then put it back together
The result will be interesting
Actually, if anyone doesn't know, this is the EMA(exponential moving average) filter:
.
OK
The problem with the Fourier series decomposition is that the problem is solved head-on, i.e. with noise
it must first be filtered and then decomposed
then apply a coefficient to each frequency or a predictor, as is customary here
and then put it back together
the result will be interesting
Yeah, welcome to machine learning.
Yeah, welcome to machine learning
Maxim, MO is DSP (digital signal processing)
So there is no point in arguing about it.
Maxim, the ME is the COC
Don't be so ridiculous
OK
The problem with the Fourier series decomposition is that the problem is solved head-on, i.e. with noise
it must first be filtered and then decomposed
then apply a coefficient to each frequency or a predictor, as is customary here
and then put it back together
the result will be interesting.
How much more can this amateur radio nonsense pestering((.
It does not work for the market, has never worked and never will.