energy: Energy of the Deep Belief Net or Restricted Bolzmann Machine.

In xrobin/DeepLearning: Deep Learning of neural networks

Description

Computes the energy of the data points in the DeepBeliefNet or RestrictedBolzmannMachine

Usage

energy(x, data, ...)

## S3 method for class 'RestrictedBolzmannMachine'
energy(x, data, drop = TRUE, ...)

## S3 method for class 'DeepBeliefNet'
energy(x, data, drop = TRUE, ...)

Arguments

x	the DeepBeliefNet or RestrictedBolzmannMachine object
data	the dataset, either as matrix or data.frame. The number of columns must match the number of nodes of the network input
...	ignored
drop	do not return additional dimensions

Value

a vector or matrix of the same size than the data (rows) giving the energy of each data point

Examples

library(mnist)
data(mnist)

# Calculate error per data point on RBM
data(pretrained.mnist)
rbm <- pretrained.mnist[[1]]
en <- energy(rbm, mnist$test$x)
head(en) # 1 value per data point

# Calculate error per data point on DBN
data(trained.mnist)
en <- energy(trained.mnist, mnist$test$x)
head(en) # 1 value per data point

# Energy is not related with reconstruction error
err <- error(trained.mnist, mnist$test$x)
## Not run: plot.mnist(predictions = cbind(err, en))

xrobin/DeepLearning documentation built on Sept. 18, 2020, 5:23 a.m.