Machine learning in trading: theory, models, practice and algo-trading - page 1018

 

Zhenya:

MT-tester cannot be customized, it's a black box too, you can't even test if it works correctly, and the speed is still orders of magnitude lower than what you can set yourself, especially with simple options for optimization. Well, how to say this ... in short you will not be taken seriously with "advisor" to any public platform, and C++ dll is used equally in Rentech and Goldmansax.

The main thing is that everything should be automatic, so I connect it and use the tester in the future when it arrives and get the result in the form of equity.


There is a contradiction here: on the one hand, you distrust the tester and, on the other hand, you rely on it.

Conclusion - you have a unique, super objective and fast tester.

I hope you will not agitate everyone to write the same tester as it is quite another, extensive topic.

In that case two variants are possible, or you will share this tester, for objective testing of models with everyone who wants or we will recognize your models ahead of time as the best of the best because we already understand that your own tester will not let you down:)

 
Ivan Negreshniy:

Contradiction - on the one hand distrust the tester, and on the other hand, on him all hope.

Conclusion - you have a unique, super objective and fast tester.

I hope you will not agitate everyone to write the same tester as it is a completely different, vast topic.

In that case two variants are possible, or you share this tester, for objective testing of models with all comers or we recognize your models ahead of time as the best of the best because we already understand that your own tester will not let you down:)

There is a tester, of course, to share it is not a pity at all is a simple enough algorithm that even surprised the very fact of the request to share it)))

There is no contradiction and there is no hope.

 
Gianni:

There is a tester, of course, to share is not a pity at all it is a simple enough algorithm that even surprised by the very fact of asking to share it)))

There is no contradiction and there is no hope.

Exactly no hope, if you do not understand that it is possible to evaluate, compare and exchange MO algorithms for trading only in the context of a single tool to measure their effectiveness - the tester, and since you only trust yours, I offered to share it.


But there is no interest in your algorithms, just like in the algorithms of your comrades on the closed forum, if your black boxes by no one knows the task and principle will only draw beautiful equity charts, even if it will be supported by poetic stories about advanced MO systems, billions of dollars and financial tycoons.

 
Aleksey Vyazmikin:

1. So I would like to know about these methods. Otherwise I invent my own bicycle again (I have already sketched the ideology), and suddenly everything has already been done before us...

2. I see. But unreasonable.

There is a lot about it in these articles.

 
mytarmailS:

Vladimir, could you tell me if there are any methods of "future selection" (or something like that) but applicable to BP? It's the algorithm which analyses BP and may exclude or add some improvements to the forecast, google couldn't help me(

We don't do anything else but time series. No?

There are a lot of methods, not only selecting, but also creating. There is a whole new section Features engineering. Look at several articles where it was discussed in details.

Good luck

 

Ivan Negreshniy:
Exactly no hope, unless you understand that it is possible to evaluate, compare and share trading MI algorithms only in the context of a single tool to measure their effectiveness - a tester, and since you only trust yours, I offered to share it.

But there is no interest in your algorithms, just as there is in the algorithms of your comrades on the closed forum, if your black boxes by some unknown task and principle will only draw beautiful equities charts, even if all this will be supported by poetic stories about advanced MO systems, billions of money and financial tycoons.

Why are you immediately offended? About the "billions of green" of course it was a joke, do not you understand, well, you're really ... I told you if you need I can share the tester algorithm, no problem. It's a long time since I used the MT-tester, and whether it is now possible to load arbitrary csv series? This is very important how else to check if it works correctly, if not then ... You know what...

I don't trade on gambling forex, I work with forts and crypto, so my mt-tester has nothing to do with it, but I know for sure that my tester works very close to real, considering the error in execution and liquidity activity in the bet. I'm more interested in forex as a data type, even having weak forecasts for major indices and indexes may help much in forecasting B-shares and even a little bit for cryptos.

 

It's funny, people write something about the tester without even bothering to look at its functionality. Not to mention that in mql you can also write your own tester and a tester within a tester

where are the "machiners" coming from on this forum and for what purpose, I'm just curious

 
Maxim Dmitrievsky:

It's funny, people are writing something about the tester without even bothering to look at its functionality. Not to mention that in mql you can also write your own tester and a tester within a tester

If so, where are the "machiners" coming from on this forum and for what purpose, I'm just curious

Maybe you're right, I may look here in 2-3 years, when the new generation of leakers comes, who knows, maybe they will be smarter, though ... I highly doubt it. Astalavista losers! Write articles and buy $50 worth of rakebacks!


 
Zhenya:

I guess you're right, look here in 2-3 years, when the new generation will come, who knows, maybe they will be smarter, though ... I highly doubt it. Astalavista losers! Write articles and rivet the $50 grail


have a nice trip )

 
Gianni:

Why are you immediately offended? About "billions of green" of course it was a joke, do not you understand, well, you're really ... I told you if you need I can share the tester algorithm, no problem. It's a long time since I used the MT-tester, and whether it is now possible to load arbitrary csv series? This is very important how else to check if it works correctly, if not then ... You know what...

I don't trade on gambling forex, I work with forts and crypto, so my mt-tester has nothing to do with it, but I know for sure that my tester works very close to real, considering the error in execution and liquidity activity in the bet. I'm more interested in forex as a data type, even having weak major and indices forecasts may help me much in forecasting B-shares and even a little bit in crypto.

Well, how can I not be offended, I thought this as a contribution to pay and almost collected...
Reason: