nnp: A supervised projection function

In AlvaroSoriano/nnproj: Data Projection by means of Neural Networks Hidden Layers

Description

This function transforms a dataset into the activations of the neurons of the hidden layer of a neural network. This is done by training a neural network and then computing the activations of the neural network for each input pattern

Usage

nnp(x, y, hidden, steps)

Arguments

x	is a data.frame that contains the inputs of the dataset
y	is a vector that contains the outpus of the dataset
hidden	is a vector that contains the number of neurons for each hidden layer
steps	is the number of iterations of the net

Examples

data <- data.frame(A=runif(10, 5.0, 7.5), B=runif(10, 1.0, 3.5), C=runif(10, 10.0, 10.5), output=runif(10,1.0, 10.0))
projection <- nnp(data[-ncol(data)], data[[ncol(data)]], hidden=2, steps=10)

AlvaroSoriano/nnproj documentation built on May 5, 2019, 4:55 a.m.