# error: Compute Reconstruction Error In xrobin/DeepLearning: Deep Learning of neural networks

## Description

Computes the reconstruction error (rmse) of the prediction of the data. `rmse` is an alias for `error`. `errorSum` sum the error over the data points.

## Usage

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23``` ```error(...) rmse(...) errorSum(...) ## S3 method for class 'DeepBeliefNet' error(x, data, ...) ## S3 method for class 'DeepBeliefNet' rmse(x, data, ...) ## S3 method for class 'DeepBeliefNet' errorSum(x, data, ...) ## S3 method for class 'RestrictedBolzmannMachine' error(x, data, ...) ## S3 method for class 'RestrictedBolzmannMachine' rmse(x, data, ...) ## S3 method for class 'RestrictedBolzmannMachine' errorSum(x, data, ...) ```

## Arguments

 `...` further arguments to the `plot` function above and to the `predict` function. `x` the `RestrictedBolzmannMachine` or `DeepBeliefNet` object `data` feature vector

## Value

the reconstruction error of the data

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23``` ```library(mnist) data(mnist) data(trained.mnist) # Calculate (reconstruction) error per data point err <- error(trained.mnist, mnist\$test\$x) length(err) # 1 value per data point # error and rmse are synonymous identical(err, rmse(trained.mnist, mnist\$test\$x)) # errorSum returns the sum sum <- errorSum(trained.mnist, mnist\$test\$x) print(sum) all.equal(sum, sum(err)) # There may be some rounding errors though, so this might not be ==: sum == sum(err) # On a RestrictedBolzmannMachine data(pretrained.mnist) rbm <- pretrained.mnist[[1]] errorSum(rbm, mnist\$test\$x) err <- error(rbm, mnist\$test\$x) identical(err, rmse(rbm, mnist\$test\$x)) ```

xrobin/DeepLearning documentation built on May 17, 2018, 3:51 a.m.