Discussion of article "Neural network: Self-optimizing Expert Advisor" - page 3

New comment
 

Afftar zhazhit :-))))

I wonder, if you write a code in Russian with transliteration and show it to Spanish audience, will they like this material?

//---------------------------------- CREA RED NEURONAL ------------------------------------------
bool creaRedNeuronal(CMultilayerPerceptronShell &objRed)
{

}

или так

//---------------------------------- sozdat' krasnyj nejron ------------------------------------------
bool sozdatKrasnyjNejron(CMultilayerPerceptronShell &objRed)
{

}

The code is not readable. I never thought that I would have to translate from Spanish to English....

 

@ Jose Miguel Soriano

how is this an EA ? it's just a script while articles headline says 

"Neural network: Self-optimizing Expert Advisor " 


i think  that its misleading since i don't see any EA attached .

 
Zhiqiang Zhu #:
The market is indeed unpredictable, just as the future is unpredictable. But, what we can do is to, through charts, perceive the market, perceive human behaviour. Because the market is driven by human behaviour. That is the purpose of our research.
Does God throw dice?
 

Test results with MLP switched off.

MlpOff

Test results with MLP turned on.

MplOn


Thank you very much for the article.

 

Three weeks of trading demo account. Lot 0.01. Leverage 1:100.

ExpertMlp

 
After updating MetaTrader, the Expert Advisor stopped learning. Who can help with fixes?
 
Alik Dmitriev #:
After updating MetaTrader, the Expert Advisor stopped learning. Who can help with fixes?

I can hack the code...

 

The problem has been solved. The error was in the preparaDatosEntra function.

In the old library it was arDatos[fila].Set(colum, valor); In the new library it is arDatos.Set(fila, colum, valor);

Now everything works as before. There are new functions, such as MLPEBaggingLBFGS. Does anyone know how to work with ensembles?

123
New comment