Amore of neural network with R language implementation

Paste the Experiment Code first
The target uses the Amore method of the neural network to train the data and then test the data

Library (amore)
X1 <-Round (runif (2000,1,2000))
X2 <-Round (runif (2000,1,2000))
X11 <-Scale (x1[1:1900])
X12 <-Scale (x2[1:1900])
x21 <-Scale (x1[1901:2000])
x22 <-Scale (x2[1901:2000])
Y1 <-x11^2+x12^2
Y2 <-x21^2+x22^2
P <-Cbind (x11,x12)
Q <-cbind (x21,x22)
Target =y1

NET <-newff (n.neurons = C (2,2,1), Learning.rate.global = 1e-2,momentum.global = 0.4
, Error.criterium = ' LMS ', Stao = Na,hidden.layer = "Tansig",
Output.layer = ' Purelin ', method = "ADAPTGDWM")
Result <-train (Net,p,target,error.criterium = "LMS", report = True,show.step = 100
, n.shows = 5)

z<-Sim (RESULT$NET,Q)
Plot (q[1:100,1],z, col= "Blue", pch= "+")
Points (q[1:100,1],y2,col= "Red", pch= "X")

NEWFF parameter Description
NEWFF (N.neurons, Learning.rate.global, Momentum.global, Error.criterium, Stao, Hidden.layer, Output.layer, method)
Functions: Creating Multilayer feedforward Neural networks
N.neurons the first element: the number of input neurons, and finally the number of output neurons, the rest are the number of different hidden layer neurons.
Learning.rate.global Learning Rate
Momentum.global Momentum Factor
Error.criterium Error measurement algorithm "LMS": error sum of squares; "lmls": Logarithmic squared difference "TAO" doesn't understand.
Stao Standard for error judgment (typically NA)
Hidden.layer hidden Layer-activation function of hidden layer "Purelin", "Tansig", "sigmoid", "Hardlim", "Custom": Custom
Output.layer hidden Layer-the activation function of the output layer
Methods for method Training
Adaptive gradient descent of "ADAPTGD"
"ADAPTGDWM" momentum adaptive gradient descent
"BATCHGD" Batch gradient descent
"BATCHGDWM" Momentum batch gradient descent
NEWFF returns an object of a multilayer feedforward neural network

Train parameter Description
Train (NET, P, T, Error.criterium, Report, Show.step, N.shows)
Function: Neural network training function

NET: Neural network training
P: Enter training set
T: Output Training Set
Error.criterium's criteria for measuring goodness of fit are generally consistent with the above net (LMS, lmls, TAO)
Report Boolean whether information is provided during training
Show.step until the training function gives the maximum value of the result
N.shows report the number of workouts

SIM (NET, P) function: Calculates the output value of a given data set neural network