Libraries: Easy Neural Network

 

Easy Neural Network:

A native implementation of neural networks in pure MQL, bundled with an easy to use interface, with easy support of saving and loading of the network configurations after training.

Author: Nelson Wanyama

 
Impressive! Thank you very much for such a beautiful work! 
 
the_rod:
Impressive! Thank you very much for such a beautiful work! 
You're welcome!
 
I can't compile the script. Is there any missing file forgotten? 
 
Aram Mohammed:
I can't compile the script. Is there any missing file forgotten? 

I can compile it. Did you create a directory with the name Nerve and placed the headerfile there?

MQL5\Include\Nerve\NerveMatrix.mqh
 

Hi!

First of all, wonderfull job, but I was trying to change the prediction, in your sample, but, it doesn't change the result.

I only did this change in your sample:     

double predict[] = {1.0, 1.0};

but the result that I'm getting is 0,0093... but it should give me 1

Do you know if I'm missing something?

 
Gabriel Ochoa:

Hi!

First of all, wonderfull job, but I was trying to change the prediction, in your sample, but, it doesn't change the result.

I only did this change in your sample:     

double predict[] = {1.0, 1.0};

but the result that I'm getting is 0,0093... but it should give me 1

Do you know if I'm missing something?

Increase the epochs. That will get the value closer, but not equal to one, unless you use the appropriate output function, like reLu or xor
 
Thanks Nelson, I will try this.
Reason: