Machine learning in trading: theory, models, practice and algo-trading - page 177

 
J.B:

I see what you mean, he's probably an asshole in this, but in the rest of the guru, not to add, not to take away.

And how do you like this: http://www.forecsys.ru/ru/site/projects/safran/ in '97? You say it has nothing to do with the case?))


I have the impression that it is not relevant to us. I can't imagine in the open access the sources of information, which would characterize the individual participants of the trades. On MICEX servers there is information about each participant, especially taking into account that there are no casual bidders at MICEX. From this point of view, the text on your link becomes clear - MICEX is obliged to fight against the manipulation of quotes on the exchange, but what does that have to do with us?
 
SanSanych Fomenko:
It seems that this does not apply to us. I can not imagine in the public domain sources of information that would characterize individual bidders. On the servers of MICEX there is information about each participant, especially taking into account that there are no casual bidders at MICEX. From this point of view, the text on your link becomes clear - MICEX must fight the manipulation of quotes on the exchange, but what does that have to do with us?

This project was made by Vorontsov, the link was about the fact that he is a practitioner from the last century. Well, about "market manipulation" it is a multi-faceted topic, just like with other crimes, a lot depends on the lawyers, you would agree with the fact that detectives investigating crimes must at least understand what and how to do at least average criminals, otherwise they have no chance to catch them.

Also you probably noticed that Konstantin Vyacheslavovich, developed his own high-level programming language for the purposes of ML and algotrading in particular, which is also not just theory and science)))

 
J.B:


Combinator:
What should have in the baggage to interest your team? Or a team like yours?

Replyed in person.

If it is not inconvenient for me too in private, whisper about what you combinatora questioned.


Thank you.

 
J.B:


Also you probably have paid attention that Konstantin Vyacheslavovich, has developed the high-level programming language for ML purposes and algotrading in particular, that too not absolutely only theory and science)))

For some reason you do not want to understand the essence of my claims: any scientific article, as well as the author of such an article belongs to science, if in the preface to this article (book, dissertation) a review of all, I emphasize ALL analogs in the field covered in the publication. And R is not a trifle.

It is the basics of science.

PS.

On the subject of Vorontsov's practical involvement in the MICEX. Judging by the focus of the 1998 order, as well as my knowledge of the bacchanalia on the stock exchange, especially on stocks not blue-chip - Vorontsov did not execute the order.

 

mytarmailS:


Guys I have as it seems to me quite a strong idea how you can extract patterns from data, I have been nurturing it for a long time and I am sure that if this method does not work but no MO will work, but need help in implementation and even in the computing power itself

If anyone is ready to join the development, come back ...

Potentially interested, depending on what exactly we are talking about. Write in person or directly here, whichever is more convenient.

 
SanSanych Fomenko:
I'm sorry, but right now there is a smell of science right out of the mouth of someone else.
 

Hello mechanics!

How's it going with artificial neural intelligence? Any progress?

 
Alexander Ivanov:

Hello mechanics!

How's it going with artificial neural intelligence? Any progress?

It's going great! Let's move slowly.....

By the way, I`m visiting this forum also just for the sake of this thread, but I`ve found nothing useful for myself or for my TS unfortunately :-(

 
I think that ifmytarmailS :told about his MEGA idea about identifying patterns, it would be interesting, but so .... I know something, but I won't tell you... Yes it is not interesting to communicate so, I have disclosed a couple of my secrets, who wanted to hear them, heard, the rest ignored and in vain .....
 

Well, I even began to wonder.

I turn to R-torturers, directly toAlexey Burnakov.

Try to save your training sample only on the days when the volume increased and OM also increased, and organize the work of the model only on such days. Then make a model for another state, when the volume fell and OM also fell and make the model work only on those days, as a result you will automatically switch between models, that will allow you to make a profit over a long time period. I'm just curious!!!!!

I'm putting the indicator file, which builds not the volumes themselves, but their difference from the previous, as well as the file where the values are taken, although it is not big, but you should try it on the minutes, it should be quite enough....

The indicator you know where to put, but the file should be put in the folder. .........\MQL4\Files\evolution-dvoid\ Although I have it all under MT4 so sorry. Please let me know the results. I'm just curious... if this scheme works or not. Because if it works, it should work under all conditions. Let's accept it as working if it will improve the results of the network.

Pys. The text file should be renamed to kcv, because the forum does not support this type