I am having a look at the DeepLearn example right now. It's nice that it is so easy to set up and experiment with the given script. But i have no idea how to do this:

I want to train the network for forcasting long or short for the next hour. (so i changed lifetime to 1, using h1)
Now it makes 1 trade every hour, for learning/training thats somehow ok. But in reallity i want,
that if long trade is open and after 1 hour the new forcast for the next hour is long again
that it just stay with the trade and does not close it (so lifetime is not the best option in this case?).
If it says short i want to revert the position. If it's not very certain (something between 0.48 and 0.52 or similar) it should go flat.

Any simple idea how to do so?

This is the part of the code of the Deep Learning example (already changed a bit for some testing):

Code:
if((vLong = adviseLong(NEURAL+BALANCED,0,
     change(1),range(1),change(24),range(24))) > 0.5)	
     enterLong();
if((vShort = adviseShort()) > 0.5) 
     enterShort();


Also i don't understand why the adviseShort() is empty? Is it automaticly the oposite of adviseLong in this case?

Last edited by easyX; 12/27/17 15:02.