Machine learning in trading: theory, models, practice and algo-trading - page 1197

 
SanSanych Fomenko:

I've written on the forum: I don't see any benefit to replacing one package with another. The whole problem is predictors and their predictive ability to a particular teacher. If this problem is solved, then at the expense of the package you can knock out a few percent of the error, but it's not worth bothering with something.

Well, so we solve it, for example, bruteforce. And someone uses pure knowledge of reality.

The problem is that there is almost nothing at all in MQL, so they start choosing who is interested in MO.

I like python because the syntax is more understandable. They are the same for P and R. In other respects the global packages are the same.

 
SanSanych Fomenko:

Python has NOTHING to do with our problems at all. Python is a pile of everything and anything useful to us, a redneck subplot with incomprehensible versions that are not compatible from top to bottom. Python's support is purely rustic, on enthusiasts. Python's acceptable package rubrication does not exist as such. Anything to find in Python is a search on forums, by enthusiasts.

as history has shown, IBM is also a poddukha which the men in the garage screwed up, I think the future is for Commodore 64 !!!

 
Igor Makanu:

as history has shown, IBM is also a dud which the guys in the garage screwed up, I think that the future is Commodore 64 !!!

SanSanitch is a tough guy... instead of fighting for R he'd better write something interesting :))

 
Maxim Dmitrievsky:

SanSanych is a tough guy... instead of rooting for R, he'd rather write something interesting :))

I read his articles, I have nothing against R, but to stop Python for the masses? - it makes no difference, more people are involved in the development, it is more probable that the"Black Swan", which may lead to a result that will go down in history ;)

 
SanSanych Fomenko:

I've written on the forum: I don't see any benefit to replacing one package with another. The whole problem is predictors and their predictive ability to a particular teacher. If this problem is solved, then at the expense of the package you can knock out a few percent of the error, but it's not worth the trouble to bother with something.

You can write a lot of things, but it is harder to prove them in practice.

This proves, that the problem is not only in predictors and probably not only in their combination with model but also in implementation of this model in a certain package.

 
Maxim Dmitrievsky:

SanSanych is a cool specialist... instead of drowning in R, he'd rather write something interesting :))

I can't finish it for half a year, but interest has dropped.


So without me, at least for now.

 
SanSanych Fomenko:

I can't finish a little bit for half a year - my interest has dropped.


So without me, at least for now.

A little bit, it always seems so... then a little bit more... :) end up taking months and years

because the final goal is not completely clear, as it should look. For example: I think I have done it and it seems possible to trade with certain tolerances, but I want better and better.

 

In my opinion python is better in the sense that there is more information in Russian, in R even I haven't found Russian speaking forum, where I could ask silly questions...

But in my opinion there is no need in bridges for model working in MT, what I need is a code, in which this model will be available for connection as a function in MQL5.

 
Aleksey Vyazmikin:

In my opinion python is better in the sense that there is more information in Russian, in R even I haven't found Russian speaking forum, where I could ask silly questions...

But in my opinion bridges are not necessary for model working in MT, what I need is a code, that will make this model available for connecting as a function in MQL5.

https://www.youtube.com/playlist?list=PLu5flfwrnSD5d02G9YJcDv30Fp5_70-sI

Основы эконометрики в R
Основы эконометрики в R
  • www.youtube.com
Задача курса — научить вас находить и оценивать зависимости в реальных данных, а также визуализировать, интерпретировать и использовать их для прогнозирования.
 

I need to communicate - video meets, I don't argue.

R has a library, but I haven't understood how to activate it without much code modification, and I haven't found someone who would do it...