h2o (version 3.10.5.3)

h2o.deepfeatures: Feature Generation via H2O Deep Learning or DeepWater Model

Description

Extract the non-linear feature from an H2O data set using an H2O deep learning model.

Usage

h2o.deepfeatures(object, data, layer)

Arguments

object

An object that represents the deep learning model to be used for feature extraction.

data

An H2OFrame object.

layer

Index (for DeepLearning, integer) or Name (for DeepWater, String) of the hidden layer to extract

Value

Returns an H2OFrame object with as many features as the number of units in the hidden layer of the specified index.

See Also

link{h2o.deeplearning} for making H2O Deep Learning models.

link{h2o.deepwater} for making H2O DeepWater models.

Examples

Run this code

library(h2o)
h2o.init()
prosPath = system.file("extdata", "prostate.csv", package = "h2o")
prostate.hex = h2o.importFile(path = prosPath)
prostate.dl = h2o.deeplearning(x = 3:9, y = 2, training_frame = prostate.hex,
                               hidden = c(100, 200), epochs = 5)
prostate.deepfeatures_layer1 = h2o.deepfeatures(prostate.dl, prostate.hex, layer = 1)
prostate.deepfeatures_layer2 = h2o.deepfeatures(prostate.dl, prostate.hex, layer = 2)
head(prostate.deepfeatures_layer1)
head(prostate.deepfeatures_layer2)

#if (h2o.deepwater.available()) {
#  prostate.dl = h2o.deepwater(x = 3:9, y = 2, backend="mxnet", training_frame = prostate.hex,
#                              hidden = c(100, 200), epochs = 5)
#  prostate.deepfeatures_layer1 =
#    h2o.deepfeatures(prostate.dl, prostate.hex, layer = "fc1_w")
#  prostate.deepfeatures_layer2 =
#    h2o.deepfeatures(prostate.dl, prostate.hex, layer = "fc2_w")
#  head(prostate.deepfeatures_layer1)
#  head(prostate.deepfeatures_layer2)
#}

Run the code above in your browser using DataCamp Workspace