Machine learning in trading: theory, models, practice and algo-trading - page 2082
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
For some reason I couldn't find any discussion of using MO for news trading on this forum. I can't find anything like that here?
If you want to estimate the profitability of MO for trading with news, you'll have to use the analytical method - look at the results of analysis and analysis. I myself did not use. The other fascinated.
The archive from here.
All the rest only parses from sites on their own. and there are no pre-grades too.
It is just that optimization by several criteria in the parameter space results in sets (Pareto) rather than individual points.
So it is not far from Lobachevsky)))) Can the balance between the motor power and comfort be considered a task of optimization? Somehow, taking into account today's understanding, I consider it as a task of optimization, as a search for the best balance. All the ergonomics is in the same place.
Let me show you a simple example...
We have a trading system based on two scales with periods of 10 and 20, trading by crossovers, classic...
The scale is a low-pass filter, the period of the scale is the controlling parameter
In this case the controlling parameter is equal to the constants 10 and 20
the problem: on a given segment of the market to get DYNAMIC (not constants) control parameters of each sledgehammer, so they would be optimal in terms of profits
this is criterion 1
criterion ¹ 2 : received dynamic parameters of control should be similar to continuous function, instead of chaotic dispersion of points, that is it is necessary to introduce the concept of smoothness of function...
How do you see the solution to such a problem?
I don't recall knowing you personally...
The problem here is not about optimization, but about the choice of the notorious predictors. For this problem the question sounds like this: "What predictors to choose for optimal MA period control?", to this, in a sense, philosophical question I do not have an answer.
Yes
Can you tell me, for dumbasses, how do I get the right periods for the mashups from the Fourier coefficients?
Determine the size of the analysis window, e.g. 101 (for symmetry). You take the data half a period ahead and need the increments (50 bars forward, so as to obtain the current bar in the middle of the window). Using these 101 bars, obtain Fourier transform (frequency spectrum). Find the maximum frequency in amplitude, there may be several frequencies, decide which one you need. Frequency is 1/period. If, for example, maximum amplitude with period is 20, then you need to remove frequencies above and below this frequency. From the cutoff frequencies you just get the periods of MA, 1/frequency
I do not recall that we were personally acquainted...
For this problem the question is: "What predictors should be chosen for optimal management of MA periods?", on this, in a sense, philosophical question, I do not have an answer.
There is no philosophy, predictors and other heresies. We need to find a function that will manage the indicator which will maximize the profit!
There is no philosophy here, predictors and other heresy, it is necessary to find a function that will manage the indicator that will give the maximum profit! The problem of maximization, the CLASSIC OPTIMIZATION!
ok.
the function controls the indicator.
what are the input variables and/or constants of this control function?
what is fed to the input of the control function?
that's it. i'm done with leading questions.
why use a wizards when you can directly use a control function in trading?
ok.
the function controls the indicator.
1) what are the input variables and/or constants of this control function?
2) what is fed to the input of the control function?
that's it. I'm out of leading questions.
1) the range from the periods of the waving to be enumerated
2) the control function is the optimal period for the machine, what can be fed to the input of the period for the machine?
karoch. you are an expert like me ilon mask.
i don't know why you should use a watchdog when you can use the open function directly in trading?
Determine the size of the window for analysis, for example 101 (for symmetry). You take the data half a period ahead, you need increments (50 bars ahead, so that the current bar is in the middle of the window). Using these 101 bars, obtain Fourier transform (frequency spectrum). Find the maximum frequency in amplitude, there may be several frequencies, decide which one you need. Frequency is 1/period. If, for example, maximum frequency with period is 20, then you need to remove frequencies above and below this frequency. Using the cutoff frequencies, you will get periods of MA, 1/frequency.
I see what you mean... But this all kind of describes an ideal non-delayed mask, and I gave two mashups for simplicity of example, there can be any set of functions...
Imagine TS of 5 indicators, we should synthesize at least 5 functions from different indicators... Fourier probably won't help here...
And in general it's not clear whether frequencies describe an ideal control...?
You're as much of an expert as I am, Elon Musk.
it depends on what you smoke
I kind of got it... But this all kind of describes an ideal non-delayed mask, and I gave two mashups for simplicity, there can be any set of functions...
Imagine TS of 5 indicators, we should synthesize at least 5 functions from different indicators... Fourier probably won't help here...
And in general it is not clear whether frequencies describe the ideal control...?
If the frequency has the maximum amplitude, it is easier to extract it from the signal and it will give the greatest profit, imagine the sum of the sines, one with an amplitude of 10, the other with an amplitude of 100.
imho, the ideal indicator is an oscillator (bandpass filter) tuned to the frequency with maximum amplitude.