Machine learning in trading: theory, models, practice and algo-trading - page 1062

New comment
 
Vizard_:

We used to be friends)))

Long time ago!!!!! I am now interested in issues of a more global nature. Of worldwide importance.... DOMINATION OOHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!!!!!


Scary?

 
FxTrader562:

Ok, I think I got it to some extent. Have you already implemented the MQL5 code and tested?

My main problem is that I am still not 100% clear as how to feed raw price data to RDF other than indicator values without fuzzy logic based on your previous article.

If you can just tell me how to feed raw price data without using fuzzy logic, then it would be great. I mean the "CalculateMamdani()" function without fuzzy logic. Otherwise, I have to wait until you publish your next article.

yes, but without gmdh... I'm not sure stiil how to do it better

double CalculateMamdani()
  {
   CopyBuffer(hnd1,0,0,1,arr1);
   NormalizeArrays(arr1);

   CopyBuffer(hnd2,0,0,1,arr2);
   NormalizeArrays(arr2);

   CopyBuffer(hnd3,0,0,1,arr3);
   NormalizeArrays(arr3);

   if(!random_policy)
     {
      vector[0]=arr1[0];
      vector[1]=arr2[0];
      vector[2]=arr3[0];

      CDForest::DFProcess(RDF,vector,RFout);
      updateNeutral.B(RFout[0]); res = RFout[0];

     }
   else
     {
      int unierr;
      updateNeutral.B(MathRandomUniform(0,1,unierr)); res = MathRandomUniform(0,1,unierr);
     }
   
   //Print(updateNeutral.B());
   firstTerm.SetAll(firstInput,arr1[0]);
   secondTerm.SetAll(secondInput,arr2[0]);
   thirdTerm.SetAll(thirdInput,arr3[0]);

   Inputs.Clear();
   Inputs.Add(firstTerm);
   Inputs.Add(secondTerm);
   Inputs.Add(thirdTerm);

   CList *FuzzResult=OurFuzzy.Calculate(Inputs);
   Output=FuzzResult.GetNodeAtIndex(0);
   double res=Output.Value();
   delete FuzzResult;

   return(res);
  }
Something like this... and delete all fuzzy logic from EA
 
Vizard_:

No way, I won't sleep today)))
Misha, archive everything you have and send it to Maksim. Let him do the digging.
There were a lot of implementations, but the number of neurons multiplied by 2,
you don't have to do that. That's one of the studies of neurobiologists...

No need to look, I'll find it.

 
Maxim Dmitrievsky:

yes, but without gmdh. I'm not sure stiil how to do it better

Something like this ... and delete all fuzzy logic from EA

Thank you very much !!!

Are you sure about this line?

res = RFout [0];

Or should it be?

res = RFout [1];

By the way of The, I of the Already tried each and every combination of Formulas, math, math the pow, etc logs folder to the folder the update this policy and the update the reward functions The of The, But Somehow Gives of IT of SO far the random results. I mean the results are not reliable always.

But I want to try random candle simulations algo which is used like in "ALPHA ZERO". Are you sure RDF can take direct prices like candle close, candle open etc. during optimization?

 
Vizard_:

I'm not going to sleep tonight)))
Misha, archive everything you have and send it to Maxim. Let him read it.
There were a lot of implementations, but the number of neurons multiplied by 2,
you don't have to do that. That's what I told one of the researches of neurobiologists...

No way... I'll do it myself. But it will take me years :-) But I'm close to the core of the optimizer. No matter what they say, 2-4 out of 10 models are generalized, the rest are not. I know that the algorithm allows to get generalized models. Now we need to improve quality of estimation and apply it in the optimizer itself to make the percentage of generalized models more than 40%. Because it will run through the models until it finds a generalized model. Or he will jump only on generalized models looking for the best one...... What are the options for evaluating generalizability that you know????

 
Maxim Dmitrievsky:

New library:

Sample EA with close prices:

Now you can learn it in tester (not optimisation) and only 1 iteration

Copy and Ill delete it, Later it will be in article

Thank you very much!!!!!!!!!!!

Copied...But it is showing some error. Should I inlclude the mt5_r library also?

 
FxTrader562:

Thank you very much!!!!!!!!!!!

Copied...But it is showing some error. Should I inlclude the mt5_r library also?

mt5_r? no such library

this library#include <RL blender 1 iteration.mqh>

must be in "include" mt5 folder

 
Maxim Dmitrievsky:

mt5_r? no such library

Sorry...Yes, I got it. I used a different name for the include file.

I will test it and let you know the results.

Mainly I am more interested in creating a random candle simulations to see if RDF can actually learn from random candle price patterns.

 
FxTrader562:

Sorry...Yes, I got it. I used a different name for the include file.

I will test it and let you know the results.

Mainly I am more interested in creating a random candle simulations to see if RDF can actually learn from random candle price patterns.

Ok, if some progress with gdmh Ill write to you

 
Maxim Dmitrievsky:

Ok, if some progress with gdmh Ill write to you

For me GDMH doesn't seem much difficult to implement if I understood it correctly...But I will look into it again

1.You calculate each polynomial by taking a for loop and getting the summation of multiplication of co-efficient and indicator value inputs like ai*xi

2.Next, feed the individual polynomial to RDF input and train it

3.Next, Calculate the optimal co-efficient using least square method

4.Next, iterate the whole process continuously during the trading period

If I understood it correctly and if I can help you in anyway, then you can write to me.

By the way, I have good sample codes for Lotoptimization() and money management() etc which can be very very helpful if you can get the accuracy and drawdown of the system to a reasonable level. The system doesn't need to be 99% accurate all the time, but drawdown and consecutive losses matter a lot.

1...105510561057105810591060106110621063106410651066106710681069...3643
New comment