Machine learning in trading: theory, models, practice and algo-trading - page 758

[Anatolii Zainchkovskii]

Maxim Dmitrievsky:

there are such fields at the object (Trf - trained forest), and only the last field is an array

maybe they contain all the information, just check it out.

//+------------------------------------------------------------------+
//| Copy|
//+------------------------------------------------------------------+
void CDecisionForest::Copy(CDecisionForest &obj)
{
//--- copy variables
m_nvars=obj.m_nvars;
m_nclasses=obj.m_nclasses;
m_ntrees=obj.m_ntrees;
m_bufsize=obj.m_bufsize;
//--- copy array
ArrayCopy(m_trees,obj.m_trees);
}

there, )))) You figured it out by yourself )))

[Maxim Dmitrievsky]

Anatolii Zainchkovskii:

//+------------------------------------------------------------------+
//| Copy|
//+------------------------------------------------------------------+
void CDecisionForest::Copy(CDecisionForest &obj)
{
//--- copy variables
m_nvars=obj.m_nvars;
m_nclasses=obj.m_nclasses;
m_ntrees=obj.m_ntrees;
m_bufsize=obj.m_bufsize;
//--- copy array
ArrayCopy(m_trees,obj.m_trees);
}

there, )))) damn, you already figured it out yourself ))))

yeah :) yeah, i'm dumb

[Anatolii Zainchkovskii]

Maxim Dmitrievsky:

Yeah :) Yeah, I'm kinda dumb.

You've stopped on the scaffolding so far? ))

[Maxim Dmitrievsky]

Anatolii Zainchkovskii:

(What about the scaffolding so far? ))

Well, I dabble with q-learning, scaffolding can be used in place of q-fi-i. They are just fast, no need to wait.

Like this I want to do just for forex :) To get rid of the damned tags for learning with a teacher

there are more introductory articles, 3 in total

http://outlace.com/rlpart3.html

[Evgeny Raspaev]

Anatolii Zainchkovskii:

I would like to know an opinion, maybe I had such an idea before. What if as an input data we represent a sequence of "photos" of the chart with all desired indicators? It is necessary to stipulate, of course, that not the photos themselves but numbered "pictures". first they should be classified by their similarity and then these classes should be numbered.

There was already a similar idea, but they used computer vision somewhere on hubra with all the libraries... I couldn't find it there. The guy was looking for patterns. If I find it, I'll send you the link.

[Anatolii Zainchkovskii]

Evgeny Raspaev:

There was already similar, only using computer vision somewhere on the hubra saw with all the libraries... Now I did not find the guy was looking for patterns. If I find it, I'll send you the link.

Don't need a link, I wonder if you got something good out of it.

[Maxim Dmitrievsky]

Evgeny Raspaev:

There was already similar, only using computer vision somewhere on the hubra saw with all the libraries... Now I did not find the guy was looking for patterns. If I find it I'll send you the link.

I have computer vision, it needs a 100-layer grid.

[Anatolii Zainchkovskii]

Maxim Dmitrievsky:

Well I dabble with q-learning, scaffolding can be used instead of q-fee. They are just fast, you don't have to wait

Like this I want to do just for forex :) To get rid of the damned tags for learning with a teacher

there are more introductory articles, 3 in total

http://outlace.com/rlpart3.html

Max, I do not know English at all, so I can not understand it unfortunately.... I know only from this resource, they explain here in Russian.)

[Maxim Dmitrievsky]

Anatolii Zainchkovskii:

Max, I do not know English at all, so I will not be able to understand unfortunately.... I know it only from this site, they explain it here in russian)

google chrome has a button - translate to russian :) translates more or less well.

[Anatolii Zainchkovskii]

Maxim Dmitrievsky:

with computer vision it's a hundred layers grid

For a start at least a rough idea, divide the picture into 4 squares or 9 squares and evaluate. (square like a pixel)