Machine learning in trading: theory, models, practice and algo-trading - page 2274
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
I'm afraid it won't work even with mathematics) Roughly speaking, because the Tsosniks have non-stationarity "not the system" that we need)
Here's a good article about nonstationarity in audio:
While stationarity can be given rigorous definitions, nonstationarity is a very broad concept, as there are infinitely many ways to depart from stationarity.
While stationarity can be given rigorous definitions, nonstationarity is a very broad concept, because there are infinitely many ways to depart from stationarity.
Slightly different, the Tsosniks in the problem have a highly noisy stationary process, we know exactly what it is, and the task is to clear the noise.
In our case, more suitable is the SB model with a kind of a mixture of different in strength and period of noisy stationary motions and even repeating in time, but we do not know exactly what to look for.
That's why it is necessary to somehow single out such periods (with a sufficiently small number of errors) and not even try to do something at other moments) Attempting to build some "universal price theory" with subsequent "breaking of the forex spine" is an obvious way to nowhere).
I agree, why break when you can roll with the flow. The main thing is to keep the balance)
A little different, the Tsosniks have a highly noisy stationary process in the problem, we know exactly what it is, and the task is to clear the noise.
This is their standard theory. Noise, by the way, is especially bound to be stationary.)
I just wrote about their attempts to get away from these standard assumptions.
In our case, the SB model is more suitable with some kind of a jumble of different in strength, period noisy stationary motions series and there are even recurring in time, but exactly what to look for is not known.
Well, yes, SB is "zero approximation," and then it's a matter of taste.)
The whole multibillion-dollar market is full of nocturnals and scalpers, and you say ... through some stationary filters, I'm not strong. We don't need a perpetual motion machine, let it change, but not immediately
https://github.com/balzer82/FFT-Python
there is more of this, but I don't understand what he did
https://github.com/snazrul1/PyRevolution/blob/master/Puzzles/DSP_For_Stock_Prices.ipynb
Maybe we are talking about logic filters? They need to sell, the more beautiful words the better. I recently coded the system, from 2011 in the monitoring, the site beautiful reports and backtests. Fortunately I found the source code. My tester shows one thing and their report is another. I started to look around and on the day I had a loss they miraculously do not have any trades.
The first link is about the load on the power grid, just to find the cycles. In econometrics they also like to show examples with explicit cycles.
I haven't dealt with the second one, but the last picture says that they used Fourier to isolate cycles and continued them into the future, the orange line consists of the same chunks, it doesn't work. Here is an indicator on this topic.
Regarding Fourier, the way I see it is this. First some decomposition is done, e.g. stl
then loops are searched through bpf
then the loops are wrapped with trade logic. (Including MO) It doesn't look complicated.
Hilbert-Huang transform.
Doesn't work
Who knows how to do scale-free recognition?
Like in this video for example...
I know that it is done through spectra (Fourier, most likely), I even know how to do it, but it seems to me that what I know is not the most effective way...
So, I am interested in how scale-free is done in scientific/industrial environment, in what tasks it is applied, where to read...
Who knows how to do scale-free recognition?
Like in this video for example...
I know that it is done through spectra (Fourier, most likely), I even know how to do it, but it seems to me that what I know is not the most effective way...
So! I am interested in how scale-free is done in scientific/industrial environment , in what tasks it is applied , where to read...
Maybe DTW?
An article on hubra about DTW application in speech recognition.
Who knows how to do scale-free recognition?
Like in this video for example...
I know that it is done through spectra (Fourier, most likely), I even know how to do it, but it seems to me that what I know is not the most efficient way...
So! I am interested in how scale-free is done in scientific/industrial environment , in what tasks it is applied , where to read...
Great sliding normalization + correlation
Grand sliding normalization + correlation