Machine learning in trading: theory, models, practice and algo-trading - page 2167
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
The weirdest thing is that this dumbass and his other radio operator took so much time, slipping some fodder in the form of a few one-dollar deals and some Mashka, proving with a clever look that it's a grail.
and how many people are involved in this pointless discussion when everyone knows that Rena is a punk.
and this mayhem starts in every thread after they come in, like ***.
Don't inflate your bubble like that or you'll burst.
You can only prove your point by real trade, not by puffing your cheeks on the forum. You're a zero so far. Even minus. Hee-hee.
You're talking about the accuracy of conversions again - yes
But what will it do for trading?
Suppose you have an ideal transformation - you decided to act as an investor - you trade on the harmonic with a large period.
You entered the market when there was an extremum of the sinusoid, then your actions must be suspended for exactly the period of this harmonic?
That's what I'm saying, it does nothing (in my opinion). We ignored really existing harmonics of even longer period.
That is what we are talking about, it does not give anything (in my opinion). We ignored really existing harmonics of even larger period.
they just will remain with suddenly disappeared edge effect
You're talking about the accuracy of conversions again - yes
but what will it do for trading?
Suppose you have an ideal transformation - you decided to act as an investor - you trade on the harmonic with a large period.
you entered the market when the sinusoidal extremum was reached, then your actions should be suspended for exactly the period of this harmonic?
the harmonics weren't invented originally to trade by them :D they were already invented by those who decided to apply it to the market, i.e. not very healthy heads
Decomposition of a series, in order to predict it, may only be used in a very limited number of cases
That is what we are talking about, it does not give anything (in my opinion). We ignored really existing harmonics of even larger period.
Even if you take into account all - all - all harmonics, what will it do?
I do not know how anyone understands the Fourier transform, I understand that the purpose of finding the basis functions is to form the correct discrete samples,
i.e. we simply get the time stamps along the X axis, which will allow us to take maximum plausible amplitude values of the useful signal, and the useful signal is quantized along the Y axis --> we get a set of discrete values at the output, which describe our signal
if we do not use part of harmonics in reverse transform, then traders think we are filtering the signal, while I believe... well this is not filtering - this is a thumb in the sky, i.e. Fourier gives nothing for filtering - discrete readings are included in our sample, then not - what is the great point here?
Imho, there are statistical methods of BP transformation, they are reasonable, and Fourier... Again about the finger - I won't repeat it
harmonics were not originally invented to trade on them :D it was already invented by those who decided to apply it to the market, i.e. not quite healthy heads
the decomposition of a series, for its prediction, can only be used in a very limited number of cases
yes!
I just finished, you got it right.
Okay, I must stop, because one moron stirred up everyone with his fantasies )))
harmonics were not originally invented to trade on them :D it was already invented by those who decided to apply it to the market, i.e. not quite healthy heads
The decomposition of a series, in order to predict it, can only be used in a very limited number of cases
Well yes, any continuous function on a segment can be decomposed into a Fourier series, but predictions outside the segment based on this decomposition are only possible if the function is periodic.
Even if you take into account all - all - all harmonics, what will it do?
I do not know how anyone understands the Fourier transform, I understand that the purpose of finding the basis functions is to form the correct discrete samples,
i.e. we simply get the time stamps along the X axis, which will allow us to take maximum plausible amplitude values of the useful signal, and the useful signal is quantized along the Y axis --> we get a set of discrete values at the output, which describe our signal
If we do not use part of harmonics in reverse transform, traders think we are filtering the signal, I believe... this is not filtering, this is a thumb in the sky, i.e. Fourier gives nothing for filtering - discrete readings are included in our sample, then not - what is the great point here?
Imho, there are statistical methods of BP transformation, they are reasonable, and Fourier... Well, again about the finger - I will not repeat
It seems to be written everywhere that Fourier is good just for periodic signals. Or close to it - with a narrow spectrum.
By the way, for MO in trading, it seems to me, the Walsh decomposition would be more suitable, but for some reason I haven't seen its mentioning on the forum.
Well, yes, any continuous function on a segment can be decomposed into a Fourier series, but predictions outside the segment based on this expansion are only possible if the function is periodic.
So? Are clever phrases hiding a positive effect?
I have my doubts. Let me open another bottle of cognac. You have a lot of fun here today.
So? Is there a positive effect behind clever phrases?
I have my doubts. I'll open another bottle of cognac. You're having a lot of fun today.
You don't need predictions.
The main thing is not to be late with the current assessment, or to the store.
Any puzzle begins with being late, no matter where or where.
;)