Description Usage Arguments Value Examples
A k-fold cross validation on the training data is performed.
1 2 3 4 | Elm.cross.valid(X.fit, Y.fit, Number.hn=10, n.blocks=5, returnmodels = FALSE,
autorangeweight=FALSE, rangeweight=1,
activation='TANH',outputBias = FALSE,
rangebias = 1)
|
X.fit |
Data matrix (numeric) containing the input values (predictors) used to train the model. |
Y.fit |
Response vector (numeric) used to train the model. |
Number.hn |
Number of units in the hidden layer. Default is |
n.blocks |
an integer specifying the desired number of cross-validation folds. Default is |
returnmodels |
whether to return the trained models. Default is |
autorangeweight |
Option whether to use the automated range used for the weights. Default is |
rangeweight |
Initial random weights on |
activation |
Activation function of the hidden layer neurons. Available functions are: 'TANH' (default) and 'SIG'. |
outputBias |
Option whether to use the bias parameter in the output layer |
rangebias |
Initial random bias on |
If returnmodels=FALSE
(default), it returns the fitted values. If returnmodels=TRUE
, it returns a list containing:.
predictionTrain |
The fitted values for the training data. |
trained.elms |
A list containing the k models used. |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | set.seed(123)
attach(wtloss)
library("scales")
#scaling the inputs/outputs
x.train <- rescale(as.matrix(wtloss$Days), to=c(-1,1))
y.train <- rescale(as.matrix(wtloss$Weight), to=c(-1,1))
#training the ELM
elm.fit.values <- rescale(Elm.cross.valid(x.train,y.train,Number.hn=5,n.blocks=5),
to= range(as.matrix(wtloss$Weight)),from=c(-1,1))
oldpar <- par(mar = c(5.1, 4.1, 4.1, 4.1))
plot(wtloss$Days, wtloss$Weight, type = "p", ylab = "Weight (kg)",
main="Weight Reduction",pch=20)
points(wtloss$Days, elm.fit.values,col=2,type='p',pch=20)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.