Man pages for YTLogos/kerasR
R Interface to the Keras Deep Learning Library

ActivationApplies an activation function to an output.
ActivityRegularizationLayer that applies an update to the cost function based input...
AdvancedActivationAdvanced activation layers
ApplicationsLoad pre-trained models
AveragePoolingAverage pooling operation
BatchNormalizationBatch normalization layer
ConstraintsApply penalties on layer parameters
ConvConvolution layers
CroppingCropping layers for 1D input (e.g. temporal sequence).
CSVLoggerCallback that streams epoch results to a csv file.
DatasetsLoad datasets
decode_predictionsDecode predictions from pre-defined imagenet networks
DenseRegular, densely-connected NN layer.
DropoutApplies Dropout to the input.
EarlyStoppingStop training when a monitored quantity has stopped...
EmbeddingEmbedding layer
expand_dimsExpand dimensions of an array
FlattenFlattens the input. Does not affect the batch size.
GaussianNoiseApply Gaussian noise layer
GlobalPoolingGlobal pooling operations
img_to_arrayConverts a PIL Image instance to a Numpy array.
InitalizersDefine the way to set the initial random weights of Keras...
keras_availableTests if keras is available on the system.
keras_checkCalled to check if keras is installed and loaded
keras_compileCompile a keras model
keras_fitFit a keras model
keras_initInitialise connection to the keras python libraries.
kerasRKeras Models in R
LayerWrapperLayer wrappers
load_imgLoad image from a file as PIL object
LoadSaveLoad and save keras models
LocallyConnectedLocally-connected layer
MaskingMasks a sequence by using a mask value to skip timesteps.
MaxPoolingMax pooling operations
ModelCheckpointSave the model after every epoch.
normalizeNormalize a Numpy array.
one_hotOne-hot encode a text into a list of word indexes
OptimizersOptimizers
pad_sequencesPad a linear sequence for an RNN input
PermutePermutes the dimensions of the input according to a given...
plot_modelPlot model architecture to a file
PredictPredict values from a keras model
preprocess_inputPreprocess input for pre-defined imagenet networks
ReduceLROnPlateauReduce learning rate when a metric has stopped improving.
RegularizersApply penalties on layer parameters
RepeatVectorRepeats the input n times.
ReshapeReshapes an output to a certain shape.
RNNRecurrent neural network layers
SequentialInitialize sequential model
TensorBoardTensorboard basic visualizations.
text_to_word_sequenceSplit a sentence into a list of words.
to_categoricalConverts a class vector (integers) to binary class matrix.
TokenizerTokenizer
UpSamplingUpSampling layers.
ZeroPaddingZero-padding layers
YTLogos/kerasR documentation built on Aug. 29, 2017, 12:42 a.m.