Machine Learning in ProOrder ProRealTime

• 11/15/2020 at 3:32 PM #150565
  Khaled

  Hi Juan, can you please help… (DAX, M1, Spread 1.2)

  Attached a programme which is a compliation of bricks from the site with a large inspiration from codes posted by you @juanj @Nicolas @robertogozzi @tob107  (Thank you).

  Appreciate your feedback on the way I’ve integrated the ML code, beacuse the result with and without the ML are the same.

  in ML, isn’t the “MaxValue” supposed to be equal to “Startingvalue+(Increment*MaxIncrement)”?

  Also, I’d like to add the possibility to chose different “Periods” of “Average” to be used later in a loop… “IF… ELSE…” seems to be limited to two choices…

  Is there a possibility to add FOR x between a and c, then mm1=Avarge[a], then mm2=Avarge[b], the mm3=Avarge[c] (a, b and c are the result of the optimization).  Then change the instruction “mmRENKO=mm” into something like:

  mmRENKO=mm1

  OR

  mmRENKO=mm2

  etc…

  Please feel free to suggest imorvment to this code.

  Many Thanks

  2 users thanked author for this post.

  murre87, ayrton78

  

  Khaled

  Participant

  Topics: 59

  Replies: 314

  Been thanked: 65 times
  11/21/2020 at 3:28 PM #151134
  ayrton78

  good work to you all

  

  ayrton78

  Participant

  Topics: 0

  Replies: 2

  Been thanked: 0 times
  08/15/2021 at 7:17 PM #175400
  eckaw

  I’m playing around with Juan J’s ML code and trying to optimise only 1 value (valueX). How do one find the optimal values for the ML code? Will backtesting all variables create a over-optimised strategy as much as when over optimising ‘normal strategies’ without ML code? Where does one start? I have already worked hard to reduce variables and my strategies now only involve 1 optimised value, which is the value I’m trying to use ML on.

  

  eckaw

  Participant

  Topics: 13

  Replies: 118

  Been thanked: 61 times
  08/15/2021 at 7:40 PM #175405
  eckaw

  What I mean, is that I get very different results even when changing MaxValue and MinValue (same for Increment/MaxIncrement/Reps). How does one find stable values and avoid overfitting when implementing this ML code?

  

  eckaw

  Participant

  Topics: 13

  Replies: 118

  Been thanked: 61 times
  01/02/2022 at 1:47 PM #184325
  Khaled

  Hey Juan,

  Happy New Year!

  Any update of your heuristic Algo code that you can share with us?

  Many thanks.

  

  Khaled

  Participant

  Topics: 59

  Replies: 314

  Been thanked: 65 times
• Author
  Posts