Machine learning in trading: theory, models, practice and algo-trading - page 975
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
There are plenty of Python+MQL5 bindings on Github. Maybe I'll create my own...
Imho, you'd better create your own.
I've made my choice, and quite consciously, because I consider Python to be an underdevelopment compared to R.
But I'm not against Python - let those who want to learn it, moreover I support such a desire to learn Python, because Python is needed to implement decision blocks in EAs, which (blocks) are difficult/complicated/impossible to implement in µl. But I am very interested in expanding such a community, especially if the testing of EAs which use Python will be really shown here. There is no such a problem with R.
https://www.mql5.com/ru/users/terentyev23 Alexey has been developing his topic for a long time.
Plus I don't understand something. Do you need some special statistics, not the usual ones, which stopped being developed in the last century and which can be done everywhere, up to Exel? For example, what could you possibly lack in this package?
https://www.statsmodels.org/stable/index.html#
The peculiarity of python packages is that they are not individual models as in R, and the whole finished library. For full functionality it is enough to install 3-4 libraries and that's it.
For example, what might you be missing from this package?
https://www.statsmodels.org/stable/index.html#
What do you need it for? You need to do an EA, not perversions with unknown operability and support.
The peculiarity of python packages is that they are not individual models as in P, but entire complete libraries.
It seems to me that you should stop speaking on matters that you don't know at all.
There are plenty of Python+MQL5 bindings on Github. Maybe I'll create my own...
Can you create your own?
For example, what might you be missing from this package?
https://www.statsmodels.org/stable/index.html#
What do you need it for? You need to do an EA, not perversions with unknown operability and support.
The peculiarity of python packages is that they are not individual models as in P, but entire complete libraries.
It seems to me that you should stop speaking out on issues that you don't know at all.
In the SINSE why do you need it? So deal with advisors, why do you need R?
It seems to me that you have already grown stupid from futile undertakings and stupid arguments together with Perervenko. You may have a good understanding of the bags, but the overall intellectual level leaves a lot to be desired.
I do not take either of them seriously, because I know that they are nonsense and toys, but you can have some fun.
The competence of a "trader" is to show at least one decent statement
The competence of the machine trader is to show the statement on the MO. The competence of a nerd and inadequate trader is to force bullshit, without knowing how to use it and earn on it.
In the Sense, why do you need it? So do your advisors, why do you need P?
And it seems to me that you have already grown stupid from your futile undertakings and stupid arguments together with Perervenko. You may have a good understanding of the bags, but the overall intellectual level leaves a lot to be desired.
I do not take either of them seriously, because I know that they are nonsense and toys, but you can have some fun.
The competence of a "trader" is to show at least one decent statement
The competence of the machine trader is to show the statement on the MO. The competence of a nerd and inadequate trader is to force bullshit, without knowing how to use it and make money on it.
The offended are the ones who are bullied!
But you are right: the measure of everything is the Expert Advisor on the real market. We are working. I will show them.
The offended are the ones who pay the price!
But you're right: the measure of everything is the Expert Advisor on the real account. Let's work. If we succeed, I will show you.
Forget it, it will not get higher than the hospital average. STATISTICS.
All this pseudoscientific stuff only confuses newbies, for example, without explaining to them that this is all ordinary overoptimization with a clever look.
it is only periodically possible to tear off pieces
Never mind, you won't get above the hospital average. STATISTICS.
pseudoscientific all this flurry only confuses newcomers, for example, not explaining to them that this is all the usual over-optimization with a clever look
only periodically you can tear off bits and pieces
The problem has been solved, and it is NOT about scientific flair.
I have predictors that do not lead to overtraining, moreover the predictive ability of predictors changes very little when the window moves outside the training sample. The prediction error of the model is less than 30%, which is the same in testing, validation, and out-of-sample training. Prediction error on 4 different model types is about the same. Prediction error is adjustable and can be reduced. Works the same on 14 currency pairs. There is evidence in R.
So far all this is confirmed in the tester.
It remains to finish writing MM and portfolio management of currency pairs - it's all on µl. I will use it in my demo version and after that I will post the test results.
Hello professors of higher science!)
How's it going, is the miracle machine ready?
wiping my hands already))
Well nayes, at least something interesting to watch. I hope this happiness will last more than half a year