[Parameter("n11", DefaultValue = 0, MinValue = -1, MaxValue = 1, Step = step)]
public double n11 { get; set; }
[Parameter("n12", DefaultValue = 0, MinValue = -1, MaxValue = 1, Step = step)]
public double n12 { get; set; }
[Parameter("n13", DefaultValue = 0, MinValue = -1, MaxValue = 1, Step = step)]
public double n13 { get; set; }
[Parameter("n14", DefaultValue = 0, MinValue = -1, MaxValue = 1, Step = step)]
public double n14 { get; set; }
I work with cAlgo for ctrader and need many input parameters like those listed above as weights for a neural network. Is there a way to generate as many input parameters as you need?
For example, when i need weights for inbetween 3 hidden layers, in which there are each 8 neurons (which would be 8x8x2 weights needed), i can just tell the programm something like "double[2,8,8]" and it creates those parameters? It would be pretty annoying when i make the neural network bigger and have to write a few thousand input parameters manually.
It would be even better it the weights could directly be stored in a 3dimensional array (between which hidden layers, from which left node, to which right node)
(It is written in c#)
Any help or tips would be appreciated!
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…