Machine learning in trading: theory, models, practice and algo-trading - page 2348

 
Maxim Dmitrievsky:

In terms of learning how to boost on quotes and drawing pictures, there is no difference.

I do not understand the whining at all. It seems that they are not good at statistics and not professors, some very specific libraries are not needed.

If a service does not require professorships, it is limited in its development. These are good times. Without knowing the basics, you can do things))))

 
Valeriy Yastremskiy:

You answered the question yourself. if a service does not require a professorship, it is limited in its development. These are great times. Without knowing the gist, you can do things))))

I am almost sure that if some professor will write an article and provide a receipt, it will be published even on R, even on the bald line

 
Maxim Dmitrievsky:

It is almost certain that if some professor will write an article and provide a crust, it will be published even on R, even on the bald line.

The main thing is to understand the essence of the problem... ))))

 
Valeriy Yastremskiy:

The main thing is to understand the essence of the problem, the problem... ))))

and some problems have no solution at all, or it's a Nobel Prize at once

and it doesn't depend on lp

 
Dmitrievsky, do you remember when I go into MO, you have to be the best?
 
Maxim Dmitrievsky:

And some problems have no solution at all, or it's a Nobel Prize at once

And for example?

 

For testing, what does it matter what language, if you know how to write in that language. In a couple of seconds, you can put it into a binary form and write it in any language. So, both languages are unnecessary for trading.

We're fighting for the pelt of the unkilled bear.

 
Andrei:

And for example?

For example, a breakdown, when the cycles stop working and the TC goes down. And you don't have enough new data for retraining

 
Maxim Dmitrievsky:

And some problems have no solution at all, or it's a Nobel Prize

and it doesn't depend on jap.

Any minimization of the sample of best solutions... it's just in the plane of the width and length of the problem field and the perimeters of their optimal solutions.

added. Yes, understanding does not depend on the language...
 
Maxim Dmitrievsky:

For example, a breakdown, when the cycles stop working and the TC goes down. And you do not have enough new data for retraining.

So the loops shouldn't work.

Reason: