Help for package deep

Type:

Package

Title:

A Neural Networks Framework

Version:

0.1.0

Author:

Brian Lee Mayer

Maintainer:

Brian <bleemayer@gmail.com>

Description:

Explore neural networks in a layer oriented way, the framework is intended to give the user total control of the internals of a net without much effort. Use classes like PerceptronLayer to create a layer of Percetron neurons, and specify how many you want. The package does all the tricky stuff internally leaving you focused in what you want. I wrote this package during a neural networks course to help me with the problem set.

License:

GPL-3

Encoding:

UTF-8

LazyData:

true

Imports:

methods

RoxygenNote:

6.1.1

NeedsCompilation:

Packaged:

2019-12-16 00:43:56 UTC; rstudio

Repository:

CRAN

Date/Publication:

2019-12-20 11:50:03 UTC

deep: A Neural Networks Framework

Description

The deep package provides classes for layers, types of neurons and the neural network as a whole.

The McCullochPitts neuron class, that implements the logic of the McCullochPitts neuron model.

Description

The McCullochPitts neuron class, that implements the logic of the McCullochPitts neuron model.

Arguments

inputs

The actual data to be fed to the nuron, this input's dimentions vary with the chosen weights dimentions.

ins

The list of vectors of inputs to the first layer in the network

outs

The list of vectors of outputs of the last layer in the network

epochs

How many rounds of training to run

tax

This is the learning rate, aka eta

maxErr

A contition to early stop the training process

Value

The computed value using the McCullochPitts model.

Vector of computed values of the same size of the last layer

Fields

ws: The matrix of weights that multiply the input vector, it can be a vector, a matrix or an array.
bias: The bias value.

Examples

# Create a dataset
dataset <- iris
dataset$Petal.Length <- NULL
dataset$Petal.Width <- NULL
dataset <- dataset[dataset$Species != "versicolor",]
dataset$Code <- as.integer(dataset$Species == "virginica")
dataset <- dataset[sample(20),]

# Create the neuron
neuron <- mcCullochPitts(c(1,1), 1)

# Train the neuron, takes a while
neuron$train(dataset[,c(1,2)], dataset[,'Code', drop=FALSE], epochs = 10)

# Check the output
neuron$output(c(1,2))

# See accuracy
dataset$Calc <- sapply(1:nrow(dataset), function(x) {
    as.integer(neuron$output(dataset[x,c(1,2)]))
})
length(which(dataset$Code==dataset$Calc))/nrow(dataset)

The McCullochPittsLayer class, that implements a layer of McCullochPitts neurons.

Description

The McCullochPittsLayer class, that implements a layer of McCullochPitts neurons.

Arguments

input

The actual data to be fed to the layer, this input's dimentions vary with the chosen n.

ins

The list of vectors of inputs to the first layer in the network

outs

The list of vectors of outputs of the last layer in the network

epochs

How many rounds of training to run

tax

This is the learning rate, aka eta

maxErr

A contition to early stop the training process

Value

The computed value using the McCullochPittsLayer model.

Vector of computed values of the same size of the last layer

Fields

n: The number of neurons to create in the layer
dims: A vector of dimensions of the inputs to the layer
neurons: A list with the internal neurons

The main NeuralNetwork class, that holds the layers.

Description

The main NeuralNetwork class, that holds the layers.

Fields

eta: The learning tax, representes the size of the weight adjustment between each epoch of training.
layers: This field is a list of the layers of the network, you can use subsetting to inspect them.

Examples

# Create a dataset
dataset <- iris
dataset$Petal.Length <- NULL
dataset$Petal.Width <- NULL
dataset <- dataset[dataset$Species != "versicolor",]
dataset$Code <- as.integer(dataset$Species == "virginica")
dataset <- dataset[sample(20),]

# Create the network
net <- neuralNet(2, perceptronLayer(1))

# Train the network, takes a while
net$train(dataset[,c(1,2), drop=FALSE], dataset[,'Code', drop=FALSE], epochs = 10)

# Check the output
net$compute(c(1,2))

# See accuracy
net$validationScore(dataset[,c(1,2), drop=FALSE], dataset[,'Code', drop=FALSE])

The Perceptron neuron class, that implements the logic of the perceptron model.

Description

The Perceptron neuron class, that implements the logic of the perceptron model.

Arguments

inputs

The actual data to be fed to the neuron, this input's dimentions vary with the chosen weights dimentions.

ins

The list of vectors of inputs to the first layer in the network

outs

The list of vectors of outputs of the last layer in the network

epochs

How many rounds of training to run

tax

This is the learning rate, aka eta

maxErr

A contition to early stop the training process

Value

The computed value using the Perceptron model.

Vector of computed values of the same size of the last layer

Fields

ws: The matrix of weights that multiply the input vector, it can be a vector, a matrix or an array.
bias: The bias value.

Examples

# Create a dataset
dataset <- iris
dataset$Petal.Length <- NULL
dataset$Petal.Width <- NULL
dataset <- dataset[dataset$Species != "versicolor",]
dataset$Code <- as.integer(dataset$Species == "virginica")
dataset <- dataset[sample(20),]

# Create the neuron
neuron <- perceptron(c(1,1), 1)

# Train the neuron, takes a while
neuron$train(dataset[,c(1,2), drop=FALSE], dataset[,'Code', drop=FALSE], epochs = 10)

# Check the output
neuron$output(c(1,2))

# See accuracy
dataset$Calc <- sapply(1:nrow(dataset), function(x) neuron$output(dataset[x,c(1,2)]))
length(which(dataset$Code==dataset$Calc))/nrow(dataset)

The PerceptronLayer class, that implements a layer of Perceptron neurons.

Description

The PerceptronLayer class, that implements a layer of Perceptron neurons.

Arguments

input

The actual data to be fed to the layer, this input's dimentions vary with the chosen n.

ins

The list of vectors of inputs to the first layer in the network

outs

The list of vectors of outputs of the last layer in the network

epochs

How many rounds of training to run

tax

This is the learning rate, aka eta

maxErr

A contition to early stop the training process

Value

The computed value using the Perceptron model.

Vector of computed values of the same size of the last layer

Fields

n: The number of neurons to create in the layer
dims: A vector of dimensions of the inputs to the layer
neurons: A list with the internal neurons