Neural network in the form of a script - page 14

New comment
 

The MA fan code has been slightly changed (simplified):


//+------------------------------------------------------------------+
//| веер МА по золотому сечению |
//+------------------------------------------------------------------+
void Get0MAEUR(int $i) {
int $a=3,$b=5,$c;
for(int $j=0;$j<=9;$j++) {
MA[1][$a][$i]=iMA(Symbol(),PERIOD_M1,$a,0,MODE_EMA,PRICE_CLOSE,$i);
MA[2][$a][$i]=iMA(Symbol(),PERIOD_M5,$a,0,MODE_EMA,PRICE_CLOSE,$i);
MA[3][$a][$i]=iMA(Symbol(),PERIOD_M15,$a,0,MODE_EMA,PRICE_CLOSE,$i);
MA[4][$a][$i]=iMA(Symbol(),PERIOD_M30,$a,0,MODE_EMA,PRICE_CLOSE,$i);
MA[5][$a][$i]=iMA(Symbol(),PERIOD_H1,$a,0,MODE_EMA,PRICE_CLOSE,$i);
MA[6][$a][$i]=iMA(Symbol(),PERIOD_H4,$a,0,MODE_EMA,PRICE_CLOSE,$i);
MA[7][$a][$i]=iMA(Symbol(),PERIOD_D1,$a,0,MODE_EMA,PRICE_CLOSE,$i);
$c=$a+$b;
$a=$b;
$b=$c;
}
}
 
edwkhan писал (а) >>

I guess I haven't stumbled across the precedent yet:).

I was referring to the 2007 championship winner.

 
YuraZ писал (а) >>

There was a code YZ_BETTER_2_3_1_1.mq4

tweaked ( I don't remember the nickname )

Gone somewhere ...

I tried that!

Alas, never got a trained network!

The question about it.

I would like to add in the same tester or who has already done, to binfiles to write, and then read them. that it was possible to run with a stop, and then anything happens.

Or maybe even better: a ready-made expert can be thrown in without inputs, you can personalise it

 
edwkhan писал (а) >>

The assessor must be trained to make decisions at any subsequent point in time, depending on the situation at hand.

In other words, the expert must learn constantly and learn to act as you would in a given situation.

So far, again, I may be wrong, the task is unsolvable.

There are too many unformalised questions.

It's a good idea . The task will not be solvable if you look at it that way in advance.

But if you think aloud, what would you like to see as the expert's behavior in this situation?

Any ideas?





 
YuraZ писал(а) >>

If you mean the script YZ_BETTER_HC_2_2.rar, I assure you it is just an experiment, and not a complete one

the grid there doesn't generate signals - it generates direction

the inputs are made by a primitive filter

no one is stopping you from adding other indicators-filters

---

Take a short stop there is also a short stop, I was just doing it to visually see the point the grid indicates a possible reversal

---

in this grid.

6 inputs fed pips distances between averages like 3-5 5-8 8-13 13-21 21-55

4-50 neurons 1st hidden layer (the number of neurons in both layers is picked up in the training)

4-50 neurons 2nd hidden layer

3 neurons output


------------- buy ---- sell -- flat

output 1 | 0.00x | 0.9xxx | 0.00x

output 2 | 0.00x | 0.00x | 0.9xx

output 3 | 0.9xx | 0.00x | 0.00x

---

on 2.6 gps training on 7 samples about 1-10 minutes

in C++ on 7 samples it takes a second to a minute to learn

---

networkers know 7 samples is too few

Yuraz,

Where do I extract your YZ BETTER files?

1...7891011121314
New comment