Machine learning in trading: theory, models, practice and algo-trading - page 2501

 
JeeyCi #:

So quote yourself instead of me to your boorishness and your rotten talk... Your shit => your shit and attribute it to yourself...

You're cutting useful links, but you're leaving yourself behind... It's sickening to see what you're turning this thread into with your inadequate responses

In the beginning responded to, if you call it your own words, inadequate posts, it includes, posts off topic, personal, and not properly stated, even if they are correct. Any response, no matter what, right, wrong led to the flood.

Made conclusions, and stopped replying.... And time and nerves are wasted less)

In general, I would like to expand on the topic of MO. Finding probability patterns is correct, but not complete. More complete is the interpretation of long or strong patterns to the real, the foundation.

In general, it is strange that we can't see here (on the forum) the ideas of manual correction of Expert Advisors with data from outside. For some reason, manual control of Expert Advisors is limited to enable and disable the EA. Either we have an unclear goal - to make data acquisition fully automatic, which is not possible, or we have to do it manually.

It is obvious to process the price series, but it is not obvious to process the news. And the political decisions to formalize it has not been decided at all yet.

I hope I'm not being too obtuse.)

 
Valeriy Yastremskiy #:

In the beginning responded to, if you call it in their own words, inadequate posts, these include, off-topic, personal, and incorrectly stated, even if they are correct. Any response, no matter what, right, wrong led to the flood.

Made conclusions, and stopped replying.... And time and nerves are wasted less)

In general, I would like to expand on the topic of MO. Finding probability patterns is a correct task, but not complete. More complete is the interpretation of long or strong patterns to the real, the foundation.

In general, it is strange that we can't see here (on the forum) the ideas of manual correction of Expert Advisors with data from outside. For some reason, manual control of Expert Advisors is limited to enable and disable the EA. Either we have an unclear goal - to make data acquisition fully automatic, which is not possible, or we have to do it manually.

It is obvious to process the price series, but it is not obvious to process the news. And the political decisions to formalize it has not been decided at all.

I hope I'm not too slow in expressing myself)

Hoping for a miracle from the MoD, I agree.
 
Renat Akhtyamov #:
What to hope for a miracle from the Ministry of Defense, I agree.
Exactly that many people have too high hopes for MO and in consequence the disappointment is deep. Neural networks it is only the tool of the mathematical device which is called to help and partially solve problems where mathematics is powerless, where there is no clear logic of algorithm of calculation inherent in mathematics as a whole, where there is no model of use of the received input data and it is necessary to find it. Otherwise, MO is just another additional tool for analysis, BUT not a magic wand that can solve your family relationships (as an example)
 
Mihail Marchukajtes #:
Neural networks are just a tool of mathematical apparatus which is called to help and partially solve problems where mathematics is powerless

Profesional in neural networks with 20 years of experience... ))))

 
mytarmailS #:

a neural network professional with 20 years of experience... ))))

Grow up SON and a lot will become clear to you, but before you grow up, get under the bench :-)
 
Mihail Marchukajtes #:
Grow up SON and a lot will become clear to you, but before you grow up, get under the bench :-)

Explain to SONY how to solve problems using mathematics which cannot be solved by mathematics.

Come on, old man, come on!)) GO ON!!!

 
Mihail Marchukajtes #:
Neural networks are just that a tool of the mathematical apparatus that is meant to help and partly to solve problems where mathematics is powerless where there is no clear logic to the algorithm calculations are inherent to mathematics in general, where there is no model for using input data and it is necessary to find it. Otherwise, MO is just another additional analysis tool, BUT not a magic wand that can solve your family relationships (as an example).

mother of God......

Who Assigned It to the Purge????

 
mytarmailS #:

So tell SONY how - mathematical apparatus solve problems in which mathematics is powerless

Come on, old man, come on))) GO!!!

Okay, I'll explain for the special dumbasses .....

There are some problems which cannot be solved with elementary mathematical algorithms or logic or the solution becomes extremely laborious and impractical. For example, it is necessary to divide the input dataset into classes. It is IMPOSSIBLE to do it with the help of mathematical formulas or any special algorithm, because the law by which these data is divided is not known and it is actually what you need to find then and use the Kohonen map or any other machine learning algorithm. There are problems with an implicit solution, such as the prediction of BP, which is not possible to do with the help of simple logic in mathematics, then they start to apply ANN. That is, neural networks are used where classical mathematics can not give an answer or is powerless in plain language, and the solution can only be found through training.

To solve such problems using elementary mathematics, it is necessary to know the solution formula, which is exactly what NS does. Finds the solution to a problem through learning....

So go and learn the basics.

You seem to be literate people, but you show your stupidity on simple examples, I'm shocked!

 
Mihail Marchukajtes #:
Exactly, many people have too high hopes for MO and the subsequent disappointment is deep. Neural networks are just a tool of mathematical apparatus which is called to help and partially solve problems where mathematics is powerless, where there is no clear logic of calculation algorithm inherent to mathematics in general, where there is no model of using received input data and it is necessary to find it. Otherwise, MOE is like another additional tool for analysis, BUT not a magic wand that can solve your relationships in the family (as an example)
Rather, it is a tool that finds the best solution on the fly, just like a tester, but in the process of trading. But it needs to be given data for analysis
 
Mihail Marchukajtes #:

Okay, I'll explain for special dull-witted people.....

There are such problems which practically cannot be solved by means of elementary mathematical algorithms or logic or the solution becomes extremely labor-consuming and unreasonable. For example, it is necessary to divide the input dataset into classes. It is IMPOSSIBLE to do it with the help of mathematical formulas or any special algorithm, because the law by which these data is divided is not known and it is actually what you need to find then and use the Kohonen map or any other machine learning algorithm. There are problems with an implicit solution, such as the prediction of BP, which is not possible to do with the help of simple logic in mathematics, then they start to apply ANN. That is, neural networks are used where classical mathematics can not give an answer or is powerless in plain language, and the solution can only be found through training.

To solve such problems using elementary mathematics, it is necessary to know the solution formula, which is exactly what NS does. Finds the solution to a problem through learning....

So go and learn the basics.

You are fucking competent people, but you show your stupidity by elementary examples, I'm shocked!

If you want to "partition the input dataset into classes", what prevents you from doing it with regular trees or ensembles of trees?