Machine learning in trading: theory, models, practice and algo-trading - page 383
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 do not even understand why codes are needed at all.) The essence of anything is faster and easier to set out in ordinary words. And we are able to write the codes ourselves.)
Yeah, I'm sitting now studying the studio of lerning machines in the azure cloud, sharing links, like look how interesting :) what other codes should be attached to it
Yeah, I'm sitting now studying the lerning machine studio in azure cloud, sharing links, like look how interesting :) what other codes should be attached to this
I never got VS 2017 again after demolishing it.( And I had such hopes for it, at least in terms of R support. And Azure too.
Something seems to me that VS 2015 won't give such a flight. By the way, some of the VS 2015 projects were not opened by the 17.
I still have not installed VS 2017, again, after its demolition.( And I had such hopes for it, at least in terms of support for R. And Azure, too.
Something seems to me that VS 2015 won't give such a flight. By the way, some of VS 2015 projects were not opened by 17.
In general I noticed that the 17-th with R somehow doesn't work stable... maybe I should just get used to it... For example, it hung up completely when installing some packages, 1 time it was when executing a script.
Before that, I had installed the last R together with Open R, but it was connecting to the wrong R and hung up when creating a project without giving any warning... I deleted R by trial and error and left only the Mycrosoft one.
It's strange that it did not re-install, it was never a problem... I think there is a solution.)
But in general, the scope of Microsoft in terms of creating a whole ecosystem ... you get a studio and almost all languages, and then the stat languages, not enough? ))
I made an example with prediction of credit risk, trained 2 models (Decision Trees and Support Vectors), and immediately compared them, convenient, you can compare any model in 2 clicks. Decision trees were better. I will start to develop it for Forex soon :) The only thing left to check the web api, how realistic it would be to use it as an Expert Advisor.
Z.U. And the best part is that here I can implement my ideas with pattern recognition via OpenCL, it's amazing. Huge possibilities.
In general I noticed that the 17-th with R somehow is not stable... maybe you just need to get used to it... For example it freezes to the core when installing some packages, 1 time it was when executing a script.
Before that I had the last R installed along with Open R, but it was connecting to the wrong R and also hung up when creating a project, without giving any warnings... by trial and error I deleted R, leaving only the Microsoft one
I have never had any problems with it... I think we can find a solution.)
But in general the scope of Microsoft is amazing in terms of creating a whole ecosystem... you get studio and almost all languages and then you get studio in the cloud... ))
Maybe this will help.
1. It is necessary to uninstall all the installed R - fortunately R works without any installations.
2. Install macrosoft R. This is different from CRAN.
3. Macrosoft R usually works with Intel library for matrix operations (from memory something like mkl). You have to look carefully at the list of installed programs to have one R and the corresponding Intel library by edition. Mismatch leads to unstable operation.
Maybe this will help.
1. It is necessary to uninstall all installed R - fortunately R works without any settings.
2. Install macrosoft R. It is different from CRAN.
3. Macrosoft R usually works with Intel library for matrix operations (from memory it looks like mkl). It is necessary to look carefully at the list of installed programs to have one R and the corresponding Intel library by edition. Mismatch leads to unstable operation.
Yes, I did, it now shows that it works with Microsoft's, with that everything is ok. The studio itself sometimes hangs when installing new libs and occasionally when running scripts, but that's not too bad. Later I will poke around more, maybe there is something with the settings, I do not know where to look for the Intel libs, I think all the latest stuff is still there
I made an example with prediction of credit risk, trained 2 models (Decision Trees and Support Vectors), and immediately compared them, convenient, you can compare any model in 2 clicks. Decision trees were better. I will start to develop it for Forex soon :) The only thing left to check the web api, how realistic it would be to use it as an Expert Advisor.
Z.U. And the best part is that here I can implement my ideas with pattern recognition via OpenCL, it's amazing. Huge possibilities.
Nice framework)))
I went to 8h trial version with their examples all clear, but is it possible to put there own csv? For example with numer.ai to try my luck?
Nice framework))))
I went to 8h trial version with their examples everything is clear, but is it possible to put there own csv? For example with numer.ai to try my luck?
There is not only 8h free, there is a free version with some restrictions. Of course you can, the section Datasets, csv just recommended as a table standard
there's not just an 8 hour free version, there's a free version with some restrictions at all. Of course you can, the section date, csv just recommended as a tabular standard
All see, thanks, but in 8h something local file download is not available, I'll try to find my microsoft account...
8h did not even try, went straight to the full free version, but it needs an account