RBFNN_train: Train a Radial Basis Function Neural Network.

In xmartin46/RBFNN: Radial Basis Function Neural Network

Description

Train a Radial Basis Function Neural Network.

Usage

RBFNN_train(X, Y, neurons, it, inputType = "original_dataset")

Arguments

X	original dataset, distance matrix or gaussian kernel matrix.
Y	the labels of X. Each observation has one label.
neurons	number of neurons in the hidden layer.
it	value used in the seed.
inputType	type of X. Available types: original_dataset (original dataset), distance_matrix (Euclidean distance matrix ) and kernel_matrix (Gaussian kernel matrix).

Value

The trained model, the centroids of the RBFNN, the standard deviations of each RBF neuron and indexes of the observations that are the centroids.

Examples

data(iris)
dataset <- iris[, 1:4]
labels <- iris[, 5]

X_train <- dataset[1:100, ]
y_train <- labels[1:100]
X_test <- dataset[101:150, ]
y_test <- labels[101:150]

neurons <- 10
seed_val <- 1
RBFNN <- RBFNN_train(as.matrix(X_train), as.matrix(y_train), neurons, seed_val)

xmartin46/RBFNN documentation built on Jan. 1, 2021, 1:43 p.m.