darchTest: Test classification network.

In maddin79/darch: Package for Deep Architectures and Restricted Boltzmann Machines

Description

Forward-propagate given data through the deep neural network and return classification accuracy using the given labels.

Usage

darchTest(darch, newdata = NULL, targets = T)

Arguments

darch	DArch instance.
newdata	New data to use, NULL to use training data.
targets	Labels for the data, NULL to use training labels (only possible when data is NULL as well).

Details

This is primarily a convenience function similar to predict.DArch with classification performance measurements instead of network output, and it returns a list of accuracy indicators (raw network error, percentage of incorrect classifications and absolute number of incorrect classifications).

Value

Vector containing error function output, percentage of incorrect classifications and absolute number of incorrect classifications.

See Also

Other darch interface functions: darchBench, darch, plot.DArch, predict.DArch, print.DArch

Examples

## Not run: 
data(iris)
model <- darch(Species ~ ., iris, retainData = T)
classificationStats <- darchTest(model)

## End(Not run)

maddin79/darch documentation built on May 21, 2019, 10:53 a.m.