Index
All Classes and Interfaces|All Packages|Constant Field Values|Serialized Form
A
- absMax(float[], float[]) - Static method in class deepnetts.tensor.Tensors
-
Returns array with max values for each position in the given input vectors.
- absMax(Tensor, Tensor) - Static method in class deepnetts.tensor.Tensors
-
Returns tensors with max value for each component of input tensors.
- absMin(float[], float[]) - Static method in class deepnetts.tensor.Tensors
- absMin(TensorBase, TensorBase) - Static method in class deepnetts.tensor.Tensors
- absPrime(float) - Static method in class deepnetts.net.layers.activation.MathFunctions
-
First derivative of the abs function.
- AbstractLayer<I,
O, - Class in deepnetts.net.layersW> -
Base class for different types of layers.
- AbstractLayer(ActivationType) - Constructor for class deepnetts.net.layers.AbstractLayer
- AbstractOptimizer - Class in deepnetts.net.train.opt
-
Skeletal implementation of the Optimizer interface to minimize effort to implement specific optimizers.
- AbstractOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.AbstractOptimizer
- AbstractScaler - Class in deepnetts.data.norm
-
Base class to simplify implementation of custom normalization procedure.
- AbstractScaler() - Constructor for class deepnetts.data.norm.AbstractScaler
- AcceleratorHandle - Interface in deepnetts.accl
- AcceleratorProvider - Interface in deepnetts.accl.spi
- AcceleratorService - Class in deepnetts.accl.spi
- AcceleratorService() - Constructor for class deepnetts.accl.spi.AcceleratorService
- AcceleratorTensorBridge - Interface in deepnetts.accl
- accept(int, int) - Method in interface deepnetts.util.RangeConsumer
- accept(Tensor) - Method in interface deepnetts.net.layers.activation.ActivationFunction
- accuracy - Variable in class deepnetts.eval.ClassificationMetrics.Stats
- activation - Variable in class deepnetts.net.layers.AbstractLayer
-
Activation function for this layer.
- ActivationFunction - Interface in deepnetts.net.layers.activation
-
Common base interface for all activation functions used in layers.
- activationType - Variable in class deepnetts.net.layers.AbstractLayer
-
Type of activation function for this layer.
- ActivationType - Enum Class in deepnetts.net.layers.activation
-
Supported types of activation functions.
- ActivationVectorizationProvider - Interface in deepnetts.accl.spi
- ActivationVectorizationService - Class in deepnetts.accl.spi
- ActivationVectorizationService() - Constructor for class deepnetts.accl.spi.ActivationVectorizationService
- ADADELTA - Enum constant in enum class deepnetts.net.train.opt.OptimizerType
- AdaDeltaOptimizer - Class in deepnetts.net.train.opt
-
Implementation of ADADELTA which is a modification of AdaGrad that uses only a limited window of previous gradients.
- AdaDeltaOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.AdaDeltaOptimizer
- ADAGRAD - Enum constant in enum class deepnetts.net.train.opt.OptimizerType
- AdaGradOptimizer - Class in deepnetts.net.train.opt
-
Implementation of ADAGRAD
Optimizer, which uses sum of squared previous gradients to adjust a global learning rate for each weight. - AdaGradOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.AdaGradOptimizer
- ADAM - Enum constant in enum class deepnetts.net.train.opt.OptimizerType
- AdamOptimizer - Class in deepnetts.net.train.opt
-
Implementation of Adam optimizer which uses an estimation of a gradient statistic (mean and variance) to adjust learning rate for each weight.
- AdamOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.AdamOptimizer
- add(float) - Method in class deepnetts.tensor.TensorBase
- add(float[], float[]) - Method in class deepnetts.eval.RootMeanSquaredError
- add(float[], float[]) - Static method in class deepnetts.tensor.Tensors
-
Adds given vector and store result in first.
- add(float, int) - Method in class deepnetts.tensor.Tensor1D
-
Adds specified value to an element at specified index position in tensor.
- add(float, int, int) - Method in class deepnetts.tensor.Tensor2D
-
Adds specified value to matrix value at position x, y
- add(float, int, int, int) - Method in class deepnetts.tensor.Tensor3D
- add(float, int, int, int, int) - Method in class deepnetts.tensor.Tensor4D
- add(Parameter) - Method in class deepnetts.automl.Parameters
- add(ExampleImage) - Method in class deepnetts.data.ImageSet
-
Adds an example image that will be used to train deep learning model to this set.
- add(Tensor1D) - Method in class deepnetts.tensor.Tensor1D
- add(Tensor1D) - Method in class deepnetts.tensor.Tensor2D
- add(TensorBase) - Method in class deepnetts.tensor.TensorBase
-
Adds specified tensor t to this tensor.
- add(String, List) - Method in class deepnetts.automl.Parameters
-
Adds a list of possible values for specified parameter name
- add(String, T...) - Method in class deepnetts.automl.Parameters
- addCBOWEmbeddingLayer(int, int) - Method in class deepnetts.net.CBOWNetwork.Builder
- addConvolutionalLayer(int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with default 3x3 filter size and default activation function.
- addConvolutionalLayer(int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with given filter size (same width and height) and default activation function.
- addConvolutionalLayer(int, int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with given width and height of convolutional filter and default type of activation function.
- addConvolutionalLayer(int, int, int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given width and height of convolutional filters, given number of channels(filters) and stride(filter step), and default type of activation function.
- addConvolutionalLayer(int, int, int, int, ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given width and height of convolutional filters, given number of channels(filters) and stride(filter step), and given type of activation function.
- addConvolutionalLayer(int, int, ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with given filter size (same width and height) and given type of activation function.
- addConvolutionalLayer(int, ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with given activation function type and default 3x3 filter size.
- addConvolutionalLayer(int, Filter) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with given settings of a convolutional filter and default type of activation function.
- addConvolutionalLayer(int, Filter, ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a convolutional layer with the given number of channels(filters), with given settings of a convolutional filter and given type of activation function.
- addFlatEmbeddingLayer(int, int) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- addFullyConnectedLayer(int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds fully connected layer with specified width and default activation function.
- addFullyConnectedLayer(int) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds to the network a fully connected layer with specified width and Relu activation function by default.
- addFullyConnectedLayer(int, ActivationType) - Method in class deepnetts.net.CBOWNetwork.Builder
- addFullyConnectedLayer(int, ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds fully connected layer with specified width and activation function.
- addFullyConnectedLayer(int, ActivationType) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds fully connected addLayer with specified width and activation function to the network.
- addFullyConnectedLayer(int, ActivationType) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- addFullyConnectedLayer(int, ActivationType) - Method in class deepnetts.net.SkipGramNetwork.Builder
- addHiddenFullyConnectedLayers(int...) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds to the network several hidden fully connected layers with specified widths and default hidden activation function by default.
- addHiddenFullyConnectedLayers(ActivationType, int...) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds fully connected hidden layers with widths given in layerWidths param and given activation function type.
- addInputLayer(int) - Method in class deepnetts.net.CBOWNetwork.Builder
- addInputLayer(int) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds input layer with the specified layerWidth (number of inputs) to the network.
- addInputLayer(int) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- addInputLayer(int) - Method in class deepnetts.net.SkipGramNetwork.Builder
- addInputLayer(int, int) - Method in class deepnetts.net.CBOWNetwork.Builder
- addInputLayer(int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Input layer with specified width and height, and 3 channels by default.
- addInputLayer(int, int) - Method in class deepnetts.net.FeedForwardNetwork.Builder
- addInputLayer(int, int) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- addInputLayer(int, int) - Method in class deepnetts.net.SkipGramNetwork.Builder
- addInputLayer(int, int, boolean) - Method in class deepnetts.net.FeedForwardNetwork.Builder
- addInputLayer(int, int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Input layer with specified width, height and number of channels (depth).
- addInputLayer(int, int, int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
- addInto(TensorBase, TensorBase) - Method in class deepnetts.tensor.TensorBase
- addLayer(AbstractLayer) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a given layer to the network.
- addLayer(AbstractLayer) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds custom layer to this network (which inherits from AbstractLayer)
- addLayer(AbstractLayer) - Method in class deepnetts.net.NeuralNetwork
-
Adds a specified layer as a next layer in the network.
- addListener(TrainingListener) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Adds training listener to this trainer.
- addMaxPoolingLayer(int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
- addMaxPoolingLayer(int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a max pooling layer with given filter size and stride(filter step).
- addMaxPoolingLayer(int, int, int) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a max pooling layer with given filter size and stride(filter step).Max pooling layer comes after convolutional layer and reduces the dimensions of the input received from the previous layer.Typically filter sizes of 2 are used, which effectively halves the dimensions of the input from the previous layer.
- addMaxPoolingLayer(Filter) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds a max pooling layer with the given filter settings.Max pooling layer comes after convolutional layer and reduces the dimensions of the input received from the previous layer.
- addOutputLayer(int, ActivationType) - Method in class deepnetts.net.CBOWNetwork.Builder
- addOutputLayer(int, ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds output layer to the neural network with specified width (number of outputs) and activation function type.
- addOutputLayer(int, ActivationType) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Adds output layer to the neural network with specified width (number of outputs) and activation function type.
- addOutputLayer(int, ActivationType) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- addOutputLayer(int, ActivationType) - Method in class deepnetts.net.SkipGramNetwork.Builder
- addOutputLayer(int, Class) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Adds output layer to the neural network with specified width (number of outputs) and layer class.
- addPatternError(TensorBase, TensorBase) - Method in class deepnetts.net.loss.BinaryCrossEntropyLoss
-
Calculates error for given actual and target patterns and adds that error to total error.
- addPatternError(TensorBase, TensorBase) - Method in class deepnetts.net.loss.CrossEntropyLoss
-
Calculates and returns outpurt error vector for specified predicted and target outputs.
- addPatternError(TensorBase, TensorBase) - Method in interface deepnetts.net.loss.LossFunction
-
Calculates pattern error for singe pattern for the specified predicted and target outputs, adds the error to total error, and returns the pattern error.
- addPatternError(TensorBase, TensorBase) - Method in class deepnetts.net.loss.MeanSquaredErrorLoss
-
Adds output error vector for the given predicted and target output vectors to total error sum and returns and error vector.
- addRegularizationSum(float) - Method in class deepnetts.net.loss.BinaryCrossEntropyLoss
- addRegularizationSum(float) - Method in class deepnetts.net.loss.CrossEntropyLoss
- addRegularizationSum(float) - Method in interface deepnetts.net.loss.LossFunction
-
Adds specified regularization sum to total loss.
- addRegularizationSum(float) - Method in class deepnetts.net.loss.MeanSquaredErrorLoss
-
Add regularization sum to total loss
- addSkipGramEmbeddingLayer(int, int) - Method in class deepnetts.net.SkipGramNetwork.Builder
- addTokenizer(Tokenizer) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- addVectorized(Tensor1D, Tensor1D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- addVectorized(Tensor2D, Tensor1D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- allocateAndCopyToDev() - Method in interface deepnetts.accl.AcceleratorTensorBridge
- apply(ActivationFunction) - Method in class deepnetts.tensor.TensorBase
- apply(Tensor) - Method in class deepnetts.util.ImagePreprocessing
- apply(Tensor3D, int) - Method in interface deepnetts.net.layers.activation.ActivationFunction
- apply(Tensor3D, int) - Method in class deepnetts.net.layers.activation.LeakyRelu
- apply(Tensor3D, int) - Method in class deepnetts.net.layers.activation.Linear
- apply(Tensor3D, int) - Method in class deepnetts.net.layers.activation.Relu
- apply(Tensor3D, int) - Method in class deepnetts.net.layers.activation.Sigmoid
- apply(Tensor3D, int) - Method in class deepnetts.net.layers.activation.SoftSign
- apply(Tensor3D, int) - Method in class deepnetts.net.layers.activation.Tanh
- apply(Tensor4D, int) - Method in interface deepnetts.net.layers.activation.ActivationFunction
- apply(Tensor4D, int) - Method in class deepnetts.net.layers.activation.LeakyRelu
- apply(Tensor4D, int) - Method in class deepnetts.net.layers.activation.Linear
- apply(Tensor4D, int) - Method in class deepnetts.net.layers.activation.Relu
- apply(Tensor4D, int) - Method in class deepnetts.net.layers.activation.Sigmoid
- apply(Tensor4D, int) - Method in class deepnetts.net.layers.activation.SoftSign
- apply(Tensor4D, int) - Method in class deepnetts.net.layers.activation.Tanh
- apply(Tensor, int, int) - Method in interface deepnetts.net.layers.activation.ActivationFunction
- apply(Tensor, int, int) - Method in class deepnetts.net.layers.activation.LeakyRelu
- apply(Tensor, int, int) - Method in class deepnetts.net.layers.activation.Linear
- apply(Tensor, int, int) - Method in class deepnetts.net.layers.activation.Relu
- apply(Tensor, int, int) - Method in class deepnetts.net.layers.activation.Sigmoid
- apply(Tensor, int, int) - Method in class deepnetts.net.layers.activation.SoftSign
- apply(Tensor, int, int) - Method in class deepnetts.net.layers.activation.Tanh
- apply(Function) - Method in class deepnetts.tensor.TensorBase
-
Applies specified function to all elements in tensor in-place.
- apply(DataSet) - Method in class deepnetts.data.norm.DecimalScaler
-
Performs normalization on the given inputs.
- apply(DataSet) - Method in class deepnetts.data.norm.MaxScaler
-
Performs normalization on the given inputs.
- apply(DataSet) - Method in class deepnetts.data.norm.MinMaxScaler
-
Performs normalization on the given inputs. x = (x-min) / (max-min)
- apply(DataSet) - Method in class deepnetts.data.norm.RangeScaler
-
Performs normalization on the given inputs. x = (x-min) / (max-min)
- apply(DataSet) - Method in class deepnetts.data.norm.Standardizer
- apply(T) - Method in interface deepnetts.data.Preprocessing
- applySoftmax(Tensor1D, float) - Method in class deepnetts.net.layers.SoftmaxOutputLayer
-
/ prva implementacija single threaded ubaci da izvrti batch u for petlji, onda paralelizuj batch
- applySoftmaxBatch(Tensor2D, Tensor2D) - Method in class deepnetts.net.layers.SoftmaxOutputLayer
- applySoftmaxBatchVectorized(Tensor2D) - Method in interface deepnetts.accl.spi.ActivationVectorizationProvider
- applySoftmaxVectorized(Tensor1D, float) - Method in interface deepnetts.accl.spi.ActivationVectorizationProvider
- applyWeightChanges() - Method in class deepnetts.net.layers.AbstractLayer
-
Applies weight changes to current weights Must be diferent for convolutional does nothing for MaxPooling Same for FullyConnected and OutputLayer
- applyWeightChanges() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- applyWeightChanges() - Method in class deepnetts.net.layers.ConvolutionalLayer
-
Apply weight changes calculated in backward pass
- applyWeightChanges() - Method in class deepnetts.net.layers.EmbeddingLayer
- applyWeightChanges() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- applyWeightChanges() - Method in class deepnetts.net.layers.FlattenLayer
- applyWeightChanges() - Method in class deepnetts.net.layers.FullyConnectedLayer
- applyWeightChanges() - Method in class deepnetts.net.layers.InputLayer
-
This method does nothing in input layer.
- applyWeightChanges() - Method in class deepnetts.net.layers.LayerNorm
- applyWeightChanges() - Method in class deepnetts.net.layers.MaxPoolingLayer
-
Does nothing for pooling layer since it does not have weights It just propagates deltas from next layer to previous through connections that had max activation in forward pass
- applyWeightChanges() - Method in class deepnetts.net.layers.OutputLayer
-
Applies weight changes after one learning iteration or batch
- applyWeightChanges() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- applyWeightChanges() - Method in class deepnetts.net.NeuralNetwork
-
Applies weight changes calculated in backward pass to all layers.
- asJson() - Method in class deepnetts.net.train.TrainingResult
- AugmentImagesTest - Class in deepnetts.util
-
just move 2(x) pix to left right up down
- AugmentImagesTest() - Constructor for class deepnetts.util.AugmentImagesTest
- average(ClassificationMetrics[]) - Static method in class deepnetts.eval.ClassificationMetrics
B
- BackpropagationTrainer - Class in deepnetts.net.train
-
Backpropagation training algorithm for feed forward and convolutional neural networks.
- BackpropagationTrainer(NeuralNetwork) - Constructor for class deepnetts.net.train.BackpropagationTrainer
-
Creates an instance of BackpropagationTrainer for the given neural network to train.
- BackpropagationTrainer(Properties) - Constructor for class deepnetts.net.train.BackpropagationTrainer
-
Creates an instance of BackpropagationTrainer with the given properties.
- backward() - Method in class deepnetts.net.layers.AbstractLayer
-
This method should implement backward pass in subclasses
- backward() - Method in interface deepnetts.net.layers.Backward
- backward() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- backward() - Method in class deepnetts.net.layers.ConvolutionalLayer
-
Backward pass for convolutional layer tweaks the weights in filters.
- backward() - Method in class deepnetts.net.layers.EmbeddingLayer
- backward() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- backward() - Method in class deepnetts.net.layers.FlattenLayer
- backward() - Method in class deepnetts.net.layers.FullyConnectedLayer
- backward() - Method in class deepnetts.net.layers.InputLayer
-
This method does nothing in the input layer, and should never be called.
- backward() - Method in interface deepnetts.net.layers.Layer
-
Performs weight parameters adjustment in backward pass during training of a neural network.
- backward() - Method in class deepnetts.net.layers.LayerNorm
- backward() - Method in class deepnetts.net.layers.MaxPoolingLayer
-
backward pass for a max(x, y) operation has a simple interpretation as only routing the gradient to the input that had the highest value in the forward pass.
- backward() - Method in class deepnetts.net.layers.OutputLayer
-
This method implements backward pass for the output layer.
- backward() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- backward() - Method in class deepnetts.net.layers.SoftmaxOutputLayer
-
Performs backward pass for this layer.
- backward() - Method in class deepnetts.net.NeuralNetwork
-
Performs a backward bass across all layers in neural network, which is the calculation of corrections for the network internal parameters (weights).
- Backward - Interface in deepnetts.net.layers
- backwardImpl - Variable in class deepnetts.net.layers.AbstractLayer
- backwardTasks - Variable in class deepnetts.net.layers.OutputLayer
- balanceDownsample(ImageSet, Map) - Static method in class deepnetts.util.ImageSetUtils
- balanceUpsample(ImageSet, Map) - Static method in class deepnetts.util.ImageSetUtils
-
Balance specified data set by adding random images of each class, to get equal number of examples for each class.
- batchMode - Variable in class deepnetts.net.layers.AbstractLayer
- batchSize - Variable in class deepnetts.net.layers.AbstractLayer
- biases - Variable in class deepnetts.net.layers.AbstractLayer
- BINARY - Enum constant in enum class deepnetts.util.ColumnType
- BinaryCrossEntropyLoss - Class in deepnetts.net.loss
-
Cross Entropy Loss is a loss function used for binary classification tasks (two classes, single output which represents probability ).
- BinaryCrossEntropyLoss(NeuralNetwork) - Constructor for class deepnetts.net.loss.BinaryCrossEntropyLoss
- bins() - Method in class deepnetts.util.Statistics.Histogram
- binWidth() - Method in class deepnetts.util.Statistics.Histogram
- BoundingBox - Class in deepnetts.util
-
Koristi se verovatno u labeleru.
- BoundingBox(int, int, int, int) - Constructor for class deepnetts.util.BoundingBox
- BoundingBox(int, int, int, int, int) - Constructor for class deepnetts.util.BoundingBox
- BoundingBox(int, int, int, int, int, float) - Constructor for class deepnetts.util.BoundingBox
- BoundingBox(int, int, int, int, int, String, float) - Constructor for class deepnetts.util.BoundingBox
- BPETokenizer - Class in deepnetts.tokenizers.bpe
- BPETokenizer(BPETokenizerParameters) - Constructor for class deepnetts.tokenizers.bpe.BPETokenizer
- BPETokenizerParameters - Class in deepnetts.tokenizers.bpe
- BPETokenizerParameters(Map, Map) - Constructor for class deepnetts.tokenizers.bpe.BPETokenizerParameters
- build() - Method in class deepnetts.net.CBOWNetwork.Builder
- build() - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Builds an instance of ConvolutionalNetwork with settings specified in this builder.
- build() - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Builds an instance of FeedForwardNetwork with settings specified in this builder.
- build() - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- build() - Method in class deepnetts.net.SkipGramNetwork.Builder
- build() - Method in class deepnetts.net.train.KFoldCrossValidation.Builder
- builder() - Static method in class deepnetts.net.CBOWNetwork
- builder() - Static method in class deepnetts.net.ConvolutionalNetwork
-
Returns a builder for the
ConvolutionalNetwork - builder() - Static method in class deepnetts.net.FeedForwardNetwork
-
Returns a builder for the
FeedForwardNetwork - builder() - Static method in class deepnetts.net.FeedForwardNetworkLM
- builder() - Static method in class deepnetts.net.SkipGramNetwork
- builder() - Static method in class deepnetts.net.train.KFoldCrossValidation
- Builder() - Constructor for class deepnetts.net.CBOWNetwork.Builder
- Builder() - Constructor for class deepnetts.net.ConvolutionalNetwork.Builder
- Builder() - Constructor for class deepnetts.net.FeedForwardNetwork.Builder
- Builder() - Constructor for class deepnetts.net.FeedForwardNetworkLM.Builder
- Builder() - Constructor for class deepnetts.net.SkipGramNetwork.Builder
- Builder() - Constructor for class deepnetts.net.train.KFoldCrossValidation.Builder
C
- calculate(float) - Method in class deepnetts.net.train.opt.AbstractOptimizer
- calculateCellsPerThread(int, int) - Static method in class deepnetts.util.DeepNettsThreadPool
- calculateDeltaBias(float, int) - Method in class deepnetts.net.train.opt.AdaDeltaOptimizer
- calculateDeltaBias(float, int) - Method in class deepnetts.net.train.opt.AdaGradOptimizer
- calculateDeltaBias(float, int) - Method in class deepnetts.net.train.opt.AdamOptimizer
- calculateDeltaBias(float, int) - Method in class deepnetts.net.train.opt.MomentumOptimizer
- calculateDeltaBias(float, int) - Method in interface deepnetts.net.train.opt.Optimizer
- calculateDeltaBias(float, int) - Method in class deepnetts.net.train.opt.RmsPropOptimizer
- calculateDeltaBias(float, int) - Method in class deepnetts.net.train.opt.SgdOptimizer
- calculateDeltaBias(Tensor1D) - Method in class deepnetts.net.train.opt.AdaDeltaOptimizer
- calculateDeltaBias(Tensor1D) - Method in class deepnetts.net.train.opt.AdaGradOptimizer
- calculateDeltaBias(Tensor1D) - Method in class deepnetts.net.train.opt.AdamOptimizer
- calculateDeltaBias(Tensor1D) - Method in class deepnetts.net.train.opt.MomentumOptimizer
- calculateDeltaBias(Tensor1D) - Method in interface deepnetts.net.train.opt.Optimizer
- calculateDeltaBias(Tensor1D) - Method in class deepnetts.net.train.opt.RmsPropOptimizer
- calculateDeltaBias(Tensor1D) - Method in class deepnetts.net.train.opt.SgdOptimizer
- calculateDeltaWeight(float, int...) - Method in class deepnetts.net.train.opt.AdaDeltaOptimizer
- calculateDeltaWeight(float, int...) - Method in class deepnetts.net.train.opt.AdaGradOptimizer
- calculateDeltaWeight(float, int...) - Method in class deepnetts.net.train.opt.AdamOptimizer
- calculateDeltaWeight(float, int...) - Method in class deepnetts.net.train.opt.MomentumOptimizer
- calculateDeltaWeight(float, int...) - Method in interface deepnetts.net.train.opt.Optimizer
-
Smoothing term to prevent division by zero if sqr grad sum becomes zero 1e-8 should be also tried https://d2l.ai/chapter_optimization/adagrad.html 1e-6 The value to use is 1e-6, 1e-8, Keras uses 1e-7 for adam
- calculateDeltaWeight(float, int...) - Method in class deepnetts.net.train.opt.RmsPropOptimizer
- calculateDeltaWeight(float, int...) - Method in class deepnetts.net.train.opt.SgdOptimizer
- calculateDeltaWeight(TensorBase) - Method in class deepnetts.net.train.opt.AdaDeltaOptimizer
- calculateDeltaWeight(TensorBase) - Method in class deepnetts.net.train.opt.AdaGradOptimizer
- calculateDeltaWeight(TensorBase) - Method in class deepnetts.net.train.opt.AdamOptimizer
- calculateDeltaWeight(TensorBase) - Method in class deepnetts.net.train.opt.MomentumOptimizer
- calculateDeltaWeight(TensorBase) - Method in interface deepnetts.net.train.opt.Optimizer
- calculateDeltaWeight(TensorBase) - Method in class deepnetts.net.train.opt.RmsPropOptimizer
- calculateDeltaWeight(TensorBase) - Method in class deepnetts.net.train.opt.SgdOptimizer
- calculateOutputError() - Method in class deepnetts.net.layers.LayerNorm
- call() - Method in class deepnetts.util.CallableIntConsumer
- call() - Method in class deepnetts.util.CallableRangeConsumer
- CallableIntConsumer - Class in deepnetts.util
-
Callable int consumer.
- CallableIntConsumer(int, int, IntConsumer) - Constructor for class deepnetts.util.CallableIntConsumer
- CallableRangeConsumer - Class in deepnetts.util
-
Callable range consumer.
- CallableRangeConsumer(int, int, RangeConsumer) - Constructor for class deepnetts.util.CallableRangeConsumer
- CBOW - Enum constant in enum class deepnetts.net.NetworkType
- CBOWEmbeddingLayer - Class in deepnetts.net.layers
- CBOWEmbeddingLayer(int, int) - Constructor for class deepnetts.net.layers.CBOWEmbeddingLayer
- CBOWNetwork - Class in deepnetts.net
- CBOWNetwork.Builder - Class in deepnetts.net
- CENTER - Enum constant in enum class deepnetts.util.ImageResize
- CenterOnWhiteBackground - Class in deepnetts.util
-
Center images on backgounds and save at target path.
- CenterOnWhiteBackground() - Constructor for class deepnetts.util.CenterOnWhiteBackground
- channelMean(int) - Method in class deepnetts.tensor.Tensor3D
- cIdx - Variable in class deepnetts.tensor.Layout
- ClassificationMetrics - Class in deepnetts.eval
-
Various metrics that tell us how good is a classifier.
- ClassificationMetrics(int, int, int, int) - Constructor for class deepnetts.eval.ClassificationMetrics
-
Constructs a new classification metrics using specified arguments.
- ClassificationMetrics(ConfusionMatrix) - Constructor for class deepnetts.eval.ClassificationMetrics
-
Constructs a new classification metrics from specified confusion matrix.
- ClassificationMetrics(ConfusionMatrix, String, int) - Constructor for class deepnetts.eval.ClassificationMetrics
-
Constructs a new classification metrics of a single class for multi class classification.
- ClassificationMetrics.Stats - Class in deepnetts.eval
-
Average values of commonly used classification metrics.
- ClassifierEvaluator - Class in deepnetts.eval
-
Evaluation method for binary and multi-class classifiers.
- ClassifierEvaluator() - Constructor for class deepnetts.eval.ClassifierEvaluator
- clone() - Method in class deepnetts.tensor.Shape
- clone() - Method in class deepnetts.tensor.Tensor1D
- clone() - Method in class deepnetts.tensor.Tensor2D
- clone() - Method in class deepnetts.tensor.TensorBase
- close() - Method in class deepnetts.net.NeuralNetwork
- COL_IDX - Static variable in interface deepnetts.net.train.opt.Optimizer
- COL_IDX - Static variable in class deepnetts.tensor.Shape
- ColorUtils - Class in deepnetts.util
- cols() - Method in class deepnetts.tensor.Tensor2D
-
Returns a number of columns in this matrix.
- cols() - Method in class deepnetts.tensor.Tensor3D
- cols() - Method in class deepnetts.tensor.Tensor4D
- columnNames - Variable in class deepnetts.data.TabularDataSet
- ColumnType - Enum Class in deepnetts.util
- ConfusionMatrix - Class in deepnetts.eval
-
Confusion matrix contains raw classifier test results.
- ConfusionMatrix(String[]) - Constructor for class deepnetts.eval.ConfusionMatrix
-
Creates a new confusion matrix for specified class labels
- containsNaN() - Method in class deepnetts.tensor.TensorBase
- CONVOLUTIONAL - Enum constant in enum class deepnetts.net.layers.LayerType
-
Performs detection of pixel patterns.
- CONVOLUTIONAL - Enum constant in enum class deepnetts.net.NetworkType
- ConvolutionalLayer - Class in deepnetts.net.layers
-
Convolutional layer performs image convolution operation on outputs of a previous layer using filters.
- ConvolutionalLayer(int, int, int) - Constructor for class deepnetts.net.layers.ConvolutionalLayer
-
Create a new instance of convolutional layer with specified number of channels filter size, default padding (filter-1)/2, and default stride stride value 1, and specified number of channels.
- ConvolutionalLayer(int, int, int, int, ActivationType) - Constructor for class deepnetts.net.layers.ConvolutionalLayer
- ConvolutionalLayer(int, Filter) - Constructor for class deepnetts.net.layers.ConvolutionalLayer
- ConvolutionalLayer(int, Filter, ActivationType) - Constructor for class deepnetts.net.layers.ConvolutionalLayer
- ConvolutionalNetwork - Class in deepnetts.net
-
Convolutional neural network is an extension of feed forward network, which can include 2D and 3D adaptive preprocessing layers (Convolutional and MaxPooling layer), which is specialized to learn to recognize features in images.
- ConvolutionalNetwork.Builder - Class in deepnetts.net
-
Builder for a convolutional neural network.
- copy() - Method in class deepnetts.tensor.Tensor1D
- copy() - Method in class deepnetts.tensor.Tensor2D
- copy() - Method in class deepnetts.tensor.Tensor3D
- copy() - Method in class deepnetts.tensor.Tensor4D
- copy() - Method in class deepnetts.tensor.TensorBase
- copy(float[], float[]) - Static method in class deepnetts.tensor.Tensors
- copy(Tensor, Tensor) - Static method in class deepnetts.tensor.Tensors
- copy2DSubChannel(int, float[]) - Method in class deepnetts.tensor.Tensor3D
-
Copies 2D channel values from this tensor into given float array buffer.
- copy3DSubChannel(int, float[]) - Method in class deepnetts.tensor.Tensor4D
- copyFrom(float[]) - Method in class deepnetts.tensor.TensorBase
- copyFrom(TensorBase) - Method in class deepnetts.tensor.TensorBase
-
Copies values from specified tensor
- copyFromGPU() - Method in class deepnetts.tensor.TensorBase
- copyOf(float[]) - Static method in class deepnetts.tensor.Tensors
- copyToDev() - Method in interface deepnetts.accl.AcceleratorTensorBridge
- copyToGPU() - Method in class deepnetts.tensor.TensorBase
- copyToHost() - Method in interface deepnetts.accl.AcceleratorTensorBridge
- correlation(float[], float[]) - Static method in class deepnetts.util.Statistics
-
Calculates and returns Pearson correlation coefficient for the given arrays.
- correlationCoefficient - Variable in class deepnetts.eval.ClassificationMetrics.Stats
- correlationMatrix(float[][]) - Static method in class deepnetts.util.Statistics
- count() - Method in class deepnetts.util.Stats
- countByClasses() - Method in class deepnetts.data.ImageSet
- countMissingValues() - Method in class deepnetts.data.TabularDataSet
- countMissingValues(int) - Method in class deepnetts.data.TabularDataSet
- create(int, int, float[]) - Static method in class deepnetts.tensor.Tensors
-
Factory method for creating tensor instance,
- create(int, int, int, float[]) - Static method in class deepnetts.tensor.Tensors
- create(int, int, int, int, float[]) - Static method in class deepnetts.tensor.Tensors
- create(ActivationType) - Static method in interface deepnetts.net.layers.activation.ActivationFunction
-
Creates and returns specified type of activation function.
- create(OptimizerType, AbstractLayer) - Static method in interface deepnetts.net.train.opt.Optimizer
-
Factory method to create different types of optimizers
- createAcceleratorBridge() - Method in class deepnetts.tensor.TensorBase
- createAcceleratorHandle(Arena) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createAcceleratorTensorBridge(TensorBase) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createBatchedDataset(TabularDataSet, int) - Static method in class deepnetts.data.DataSets
- createBatchedDatasetGPU(ImageSet, int) - Static method in class deepnetts.data.DataSets
- createBatchedDatasetGPU(TabularDataSet, int) - Static method in class deepnetts.data.DataSets
- createColsCache() - Method in class deepnetts.tensor.Tensor2D
- createConvolutionalBackwardAcc(AcceleratorHandle, ConvolutionalLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createConvolutionalForwardAcc(AcceleratorHandle, ConvolutionalLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createConvolutionalNetworkFromJson(JSONObject) - Static method in class deepnetts.util.FileIO
- createDataSetFromRawImages(String, String, int, int, boolean) - Static method in class deepnetts.util.ImageSetUtils
-
Creates image data set from raw images by resizing and randomly croping to target dimensions.
- createFeedForwardNetworkFromJson(JSONObject) - Static method in class deepnetts.util.FileIO
- createFilterCache() - Method in class deepnetts.net.layers.ConvolutionalLayer
- createFlattenBackwardAcc(AcceleratorHandle, FlattenLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createFlattenForwardAcc(AcceleratorHandle, FlattenLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createFrom(ConfusionMatrix) - Static method in class deepnetts.eval.ClassificationMetrics
-
Creates classification metrics from the given confusion matrix.
- createFromFile(File) - Static method in class deepnetts.util.FileIO
- createFromFile(File, Class) - Static method in class deepnetts.util.FileIO
- createFromFile(String, Class) - Static method in class deepnetts.util.FileIO
- createFromJson(File) - Static method in class deepnetts.util.FileIO
- createFromJson(String) - Static method in class deepnetts.util.FileIO
- createFromJson(JSONObject) - Static method in class deepnetts.util.FileIO
- createFullyConnectedBackwardAcc(AcceleratorHandle, FullyConnectedLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createFullyConnectedForwardAcc(AcceleratorHandle, FullyConnectedLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createImageIndex(String) - Static method in class deepnetts.util.ImageSetUtils
-
Creates an index file with list of all files in all subdirectories and write them into single index.txt file.
- CreateImageIndex - Class in deepnetts.util
- CreateImageIndex() - Constructor for class deepnetts.util.CreateImageIndex
- createIndexFile(HashMap, String, boolean) - Static method in class deepnetts.util.ImageUtils
- createInputFromPixels(BufferedImage, int) - Method in class deepnetts.data.ExampleImage
- createLabelsIndex(String) - Static method in class deepnetts.util.ImageSetUtils
-
Creates a file with list of labels/categories from subdirectories at the specified given path.
- CreateLabelsIndex - Class in deepnetts.util
- CreateLabelsIndex() - Constructor for class deepnetts.util.CreateLabelsIndex
- createMaxpoolingBackwardAcc(AcceleratorHandle, MaxPoolingLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createMaxpoolingForwardAcc(AcceleratorHandle, MaxPoolingLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createNeuralNetwork(Properties) - Method in class deepnetts.automl.FeedForwardNetworkFactory
- createNeuralNetwork(Properties) - Method in interface deepnetts.automl.NetworkFactory
-
Creates and returns a neural network of specified type
- createOutputBackwardAcc(AcceleratorHandle, OutputLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createOutputForwardAcc(AcceleratorHandle, OutputLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createRandomlyCroppedImages(String, String, int, int, int) - Static method in class deepnetts.util.ImageSetUtils
- createRowsAndColsCache() - Method in class deepnetts.tensor.Tensor2D
- createRowsCache() - Method in class deepnetts.tensor.Tensor2D
- createSearchSpace() - Method in class deepnetts.automl.Parameters
-
Returns parameters search space - all possible combination of all given parameter values.
- createSoftmaxOutputBackwardAcc(AcceleratorHandle, SoftmaxOutputLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createSoftmaxOutputForwardAcc(AcceleratorHandle, SoftmaxOutputLayer) - Method in interface deepnetts.accl.spi.AcceleratorProvider
- createsTrainingSnaphots() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Returns true if network creates training snapshots, false otherwise.
- createSubSampledImageIndex(String, String, int, boolean) - Static method in class deepnetts.util.ImageSetUtils
-
Copies specified number of samples of each class from
- cropAtCornersAndCenter(BufferedImage, int, int) - Static method in class deepnetts.util.ImageUtils
-
Returns five crops of a given image: four from each corner and one at center
- CROSS_ENTROPY - Enum constant in enum class deepnetts.net.loss.LossType
-
Cross Entropy Loss, used for classificaton tasks, implemented by
CrossEntropyLoss - CrossEntropyLoss - Class in deepnetts.net.loss
-
Average Cross Entropy Loss function commonly used for multi class classification problems.
- CrossEntropyLoss(NeuralNetwork) - Constructor for class deepnetts.net.loss.CrossEntropyLoss
- CsvFormat - Class in deepnetts.util
-
CSV file format options: delimiter, header(column names) and column types.
- CsvFormat() - Constructor for class deepnetts.util.CsvFormat
- cudaHandles - Variable in class deepnetts.net.layers.AbstractLayer
- cudaHandles - Variable in class deepnetts.net.NeuralNetwork
D
- DataSetItem - Class in deepnetts.data
-
Represents a basic data set item (single row) with input tensor and target vector in a data set.
- DataSetItem(float[], float[]) - Constructor for class deepnetts.data.DataSetItem
- DataSetItem(TensorBase, TensorBase) - Constructor for class deepnetts.data.DataSetItem
- DataSets - Class in deepnetts.data
-
Data set utility methods ans constants.
- DataSets() - Constructor for class deepnetts.data.DataSets
- DEBUG - Enum constant in enum class deepnetts.net.Mode
- Debugger - Class in deepnetts.util
- Debugger() - Constructor for class deepnetts.util.Debugger
- DECIMAL - Enum constant in enum class deepnetts.util.ColumnType
- DecimalScaler - Class in deepnetts.data.norm
-
Decimal scale normalization for the given data set.
- DecimalScaler(DataSet) - Constructor for class deepnetts.data.norm.DecimalScaler
-
Creates a new instance of max normalizer initialized to max values in given data set.
- decode(List) - Method in class deepnetts.tokenizers.bpe.BPETokenizer
- decode(List) - Method in interface deepnetts.tokenizers.Tokenizer
- decode(List) - Method in class deepnetts.tokenizers.WordTokenizer
- DeepNetts - Class in deepnetts.core
-
Global configuration and settings for Deep Netts Engine.
- deepnetts.accl - package deepnetts.accl
- deepnetts.accl.spi - package deepnetts.accl.spi
- deepnetts.automl - package deepnetts.automl
-
Support for automatically building deep learning models using hyper-parameter search.
- deepnetts.core - package deepnetts.core
-
Core engine configuration and settings and runtime properties.
- deepnetts.data - package deepnetts.data
-
Data structures to store example data used for building machine learning models.
- deepnetts.data.norm - package deepnetts.data.norm
-
Data normalization methods, used to scale data to specific range, in order to make them suitable for use by a neural network.
- deepnetts.eval - package deepnetts.eval
-
Evaluation procedures for machine learning models, used to estimate how good models are performing when given new data that (that was not used for training).
- deepnetts.net - package deepnetts.net
-
Neural network architectures with their corresponding builders.
- deepnetts.net.layers - package deepnetts.net.layers
-
Neural network layers, which are main building blocks of a neural network.
- deepnetts.net.layers.activation - package deepnetts.net.layers.activation
-
Activation functions for neural network layers.
- deepnetts.net.loss - package deepnetts.net.loss
-
Commonly used loss functions, which are used to calculate error during the training as a difference between predicted and target output.
- deepnetts.net.train - package deepnetts.net.train
-
Training algorithms and related utilities.
- deepnetts.net.train.opt - package deepnetts.net.train.opt
-
Optimization methods used by training algorithm.
- deepnetts.net.weights - package deepnetts.net.weights
-
Weights randomization techniques, used for initializing layer's internal parameters.
- deepnetts.tensor - package deepnetts.tensor
- deepnetts.tokenizers - package deepnetts.tokenizers
- deepnetts.tokenizers.bpe - package deepnetts.tokenizers.bpe
- deepnetts.util - package deepnetts.util
-
Various utility classes including Tensor, image operations, multithreading, exceptions etc.
- DeepNettsException - Exception Class in deepnetts.util
- DeepNettsException() - Constructor for exception class deepnetts.util.DeepNettsException
-
Creates a new instance of
JDeepNettsExceptionwithout detail message. - DeepNettsException(String) - Constructor for exception class deepnetts.util.DeepNettsException
-
Constructs an instance of
JDeepNettsExceptionwith the specified detail message. - DeepNettsException(String, Throwable) - Constructor for exception class deepnetts.util.DeepNettsException
- DeepNettsException(Throwable) - Constructor for exception class deepnetts.util.DeepNettsException
- DeepNettsLogManager - Class in deepnetts.util
- DeepNettsLogManager() - Constructor for class deepnetts.util.DeepNettsLogManager
- DeepNettsRunner - Class in deepnetts.util
- DeepNettsRunner() - Constructor for class deepnetts.util.DeepNettsRunner
- DeepNettsThreadPool - Class in deepnetts.util
-
Dedicated thread pool for Deep Netts Engine.
- DeepNettsThreadPool() - Constructor for class deepnetts.util.DeepNettsThreadPool
- DeepNettsThreadPool(int) - Constructor for class deepnetts.util.DeepNettsThreadPool
- defaultProvider() - Static method in class deepnetts.accl.spi.AcceleratorService
- defaultProvider() - Static method in class deepnetts.accl.spi.ActivationVectorizationService
- defaultProvider() - Static method in class deepnetts.accl.spi.TensorVectorizationService
- DeleteAugmentImages - Class in deepnetts.util
-
just move 2(x) pix to left right up down
- DeleteAugmentImages() - Constructor for class deepnetts.util.DeleteAugmentImages
- DELIMITER_COMMA - Static variable in class deepnetts.data.DataSets
- DELIMITER_SEMICOLON - Static variable in class deepnetts.data.DataSets
- DELIMITER_SPACE - Static variable in class deepnetts.data.DataSets
- DELIMITER_TAB - Static variable in class deepnetts.data.DataSets
- deltaBiases - Variable in class deepnetts.net.layers.AbstractLayer
- deltaBiases - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- deltaInputs - Variable in class deepnetts.net.layers.LayerNorm
- deltas - Variable in class deepnetts.net.layers.AbstractLayer
-
Deltas used for learning.
- deltas - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- deltaWeights - Variable in class deepnetts.net.layers.AbstractLayer
-
Weight changes for current and previous iteration.
- deltaWeights - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- deNormalizeInputs(TensorBase) - Method in class deepnetts.data.norm.MaxScaler
- deNormalizeOutputs(TensorBase) - Method in class deepnetts.data.norm.MaxScaler
-
De-normalize given output vector in-place.
- depth - Variable in class deepnetts.net.layers.AbstractLayer
- depth() - Method in class deepnetts.tensor.Tensor3D
- depth() - Method in class deepnetts.tensor.Tensor4D
- DEPTH_IDX - Static variable in class deepnetts.tensor.Shape
- descriptiveStatistics(float[]) - Static method in class deepnetts.util.Statistics
- detectCsvFormat(String) - Static method in class deepnetts.data.DataSets
- div(float) - Method in class deepnetts.tensor.TensorBase
-
Divide all values in this tensor with specified value.
- div(float[]) - Method in class deepnetts.tensor.TensorBase
- div(float[], float) - Static method in class deepnetts.tensor.Tensors
- div(float[], float[]) - Static method in class deepnetts.tensor.Tensors
- div(TensorBase) - Method in class deepnetts.tensor.TensorBase
-
Element-wise divison with specified tensor.
- divChannel(float, int) - Method in class deepnetts.tensor.Tensor3D
- DIVIDE_AND_CONQUER - Static variable in class deepnetts.automl.HyperParameterSearch
- dotProduct(Tensor1D, Tensor1D, Tensor1D) - Static method in class deepnetts.tensor.Tensors
- dotProduct(Tensor2D, Tensor1D, Tensor1D) - Static method in class deepnetts.tensor.Tensors
-
Dot product matrix vector multiplication C = A .
- dotProduct(Tensor2D, Tensor2D, Tensor2D) - Static method in class deepnetts.tensor.Tensors
- dotProductBuffered(Tensor2D, Tensor1D, Tensor1D) - Static method in class deepnetts.tensor.Tensors
-
Matrix vector dot product.
- dotProductBuffered(Tensor2D, Tensor2D, Tensor2D) - Static method in class deepnetts.tensor.Tensors
E
- EmbeddingLayer - Class in deepnetts.net.layers
- EmbeddingLayer(int, int) - Constructor for class deepnetts.net.layers.EmbeddingLayer
- encode(String) - Method in class deepnetts.tokenizers.bpe.BPETokenizer
- encode(String) - Method in interface deepnetts.tokenizers.Tokenizer
- encode(String) - Method in class deepnetts.tokenizers.WordTokenizer
- ENUM - Enum constant in enum class deepnetts.util.ColumnType
- EPOCH_FINISHED - Enum constant in enum class deepnetts.net.train.TrainingEvent.Type
-
Specifies a training epoch finished event
- EPOCH_FINISHED - Static variable in class deepnetts.net.train.TrainingEvent
-
Specifies that training epoch has finished.
- EPOCH_STARTED - Enum constant in enum class deepnetts.net.train.TrainingEvent.Type
-
Specifies a training epoch started event
- EPOCH_STARTED - Static variable in class deepnetts.net.train.TrainingEvent
-
Specifies that training epoch has started.
- equals(TensorBase, float) - Method in class deepnetts.tensor.TensorBase
- equals(Object) - Method in class deepnetts.automl.Parameter
- equals(Object) - Method in class deepnetts.automl.Parameters.ParameterCombination
- equals(Object) - Method in class deepnetts.tensor.Shape
- equals(Object) - Method in class deepnetts.tensor.TensorBase
- equals(Object) - Method in record class deepnetts.tokenizers.bpe.Pair
-
Indicates whether some other object is "equal to" this one.
- equalsName(String) - Method in enum class deepnetts.net.layers.LayerType
- equalsName(String) - Method in enum class deepnetts.net.loss.LossType
- equalsName(String) - Method in enum class deepnetts.net.NetworkType
- evaluate(NeuralNetwork, DataSet) - Method in class deepnetts.eval.ClassifierEvaluator
-
Performs classifier evaluation and returns classification performance metrics.
- evaluate(NeuralNetwork, DataSet) - Method in class deepnetts.eval.RegresionEvaluator
- evaluateClassifier(NeuralNetwork, DataSet) - Static method in class deepnetts.eval.Evaluators
- evaluateRegressor(NeuralNetwork, DataSet) - Static method in class deepnetts.eval.Evaluators
-
Evaluates specified neural network with test set, as a regression model and returns basic regression evauation metrics.
- evaluator(Evaluator) - Method in class deepnetts.automl.HyperParameterSearch
- evaluator(Evaluator) - Method in class deepnetts.net.train.KFoldCrossValidation.Builder
- Evaluators - Class in deepnetts.eval
-
Utility methods for evaluating machine learning models.
- ExampleImage - Class in deepnetts.data
-
Example image to train a deep learning model.
- ExampleImage(BufferedImage) - Constructor for class deepnetts.data.ExampleImage
- ExampleImage(BufferedImage, String) - Constructor for class deepnetts.data.ExampleImage
- ExampleImage(BufferedImage, String, int) - Constructor for class deepnetts.data.ExampleImage
- ExampleImage(BufferedImage, String, int, int) - Constructor for class deepnetts.data.ExampleImage
- ExampleImage(File, BufferedImage, String) - Constructor for class deepnetts.data.ExampleImage
- ExampleImage(File, String) - Constructor for class deepnetts.data.ExampleImage
-
Creates an instance of new example image with specified image and label Loads image from specified file and creates matrix structures with color information
- ExampleImage(String, String) - Constructor for class deepnetts.data.ExampleImage
F
- FALSE_NEGATIVE - Static variable in class deepnetts.eval.ConfusionMatrix
-
A label for items falsely classified as negative, which are actually positive.
- FALSE_POSITIVE - Static variable in class deepnetts.eval.ConfusionMatrix
-
A label for items falsely classified as positive, which are actually negative.
- FEEDFORWARD - Enum constant in enum class deepnetts.net.NetworkType
- FEEDFORWARDLM - Enum constant in enum class deepnetts.net.NetworkType
- FeedForwardNetwork - Class in deepnetts.net
-
Feed forward neural network architecture, also known as Multi Layer Perceptron.
- FeedForwardNetwork.Builder - Class in deepnetts.net
-
Builder of a
FeedForwardNetworkinstance. - FeedForwardNetworkFactory - Class in deepnetts.automl
-
Factory for FeedForwardNetwork.
- FeedForwardNetworkFactory() - Constructor for class deepnetts.automl.FeedForwardNetworkFactory
- FeedForwardNetworkLM - Class in deepnetts.net
- FeedForwardNetworkLM.Builder - Class in deepnetts.net
- FILE_TRAINING_PROP - Static variable in class deepnetts.util.DeepNettsRunner
- FileIO - Class in deepnetts.util
-
File utilities for saving and loading neural networks.
- fill(float) - Method in class deepnetts.tensor.TensorBase
-
Fills the entire tensor with specified value.
- fillFourthDim(Tensor4D, int, float) - Static method in class deepnetts.tensor.Tensors
- Filter - Class in deepnetts.net.layers
-
Settings of a convolutional filter which is used to learn to detect pixel patterns.
- Filter(int) - Constructor for class deepnetts.net.layers.Filter
- Filter(int, int) - Constructor for class deepnetts.net.layers.Filter
- Filters - Class in deepnetts.net.layers
- Filters() - Constructor for class deepnetts.net.layers.Filters
- FlatEmbeddingLayer - Class in deepnetts.net.layers
- FlatEmbeddingLayer(int, int) - Constructor for class deepnetts.net.layers.FlatEmbeddingLayer
- FlattenLayer - Class in deepnetts.net.layers
-
Transforms outputs from previous 3D layer into a flatten 1D tensor in forward pass, Backward pass propagates weighted errors/deltas from the next fully connected layer.
- FlattenLayer() - Constructor for class deepnetts.net.layers.FlattenLayer
- flipHorizontal(BufferedImage) - Static method in class deepnetts.util.ImageUtils
- flipVertical(BufferedImage) - Static method in class deepnetts.util.ImageUtils
- forward() - Method in class deepnetts.net.layers.AbstractLayer
-
This method should implement forward pass in subclasses
- forward() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- forward() - Method in class deepnetts.net.layers.ConvolutionalLayer
-
Forward pass for convolutional layer.
- forward() - Method in class deepnetts.net.layers.EmbeddingLayer
- forward() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- forward() - Method in class deepnetts.net.layers.FlattenLayer
- forward() - Method in interface deepnetts.net.layers.Forward
- forward() - Method in class deepnetts.net.layers.FullyConnectedLayer
- forward() - Method in class deepnetts.net.layers.InputLayer
-
This method does nothing in the input layer, and should never be called.
- forward() - Method in interface deepnetts.net.layers.Layer
-
Performs calculation of layer outputs in forward pass of a neural network .
- forward() - Method in class deepnetts.net.layers.LayerNorm
- forward() - Method in class deepnetts.net.layers.MaxPoolingLayer
-
Max pooling forward pass outputs the max value for each filter position.
- forward() - Method in class deepnetts.net.layers.OutputLayer
-
This method implements forward pass for the output layer.
- forward() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- forward() - Method in class deepnetts.net.layers.SoftmaxOutputLayer
-
Forward pass for the output layer, calculates layer outputs using softmax function.
- forward() - Method in class deepnetts.net.NeuralNetwork
-
Performs a forward pass - calculations of all layers in the network for the network's current input.
- Forward - Interface in deepnetts.net.layers
- forwardImpl - Variable in class deepnetts.net.layers.AbstractLayer
- forwardTasks - Variable in class deepnetts.net.layers.OutputLayer
- fourthDim() - Method in class deepnetts.tensor.Tensor4D
- fromFile(String) - Static method in class deepnetts.tensor.Tensor1D
- fromFile(String) - Static method in class deepnetts.tensor.Tensor2D
- fromFile(String) - Static method in class deepnetts.tensor.Tensor3D
- fromFile(String) - Static method in class deepnetts.tensor.Tensor4D
- fromFile(String) - Static method in class deepnetts.tensor.Tensors
- fScore - Variable in class deepnetts.eval.ClassificationMetrics.Stats
- FULLY_CONNECTED - Enum constant in enum class deepnetts.net.layers.LayerType
-
Looks for patterns in inputs.
- FullyConnectedLayer - Class in deepnetts.net.layers
-
Fully connected layer is used as a hidden layer in a neural network, and it has a single row of units/nodes/neurons connected to all neurons in previous and next layer.
- FullyConnectedLayer(int) - Constructor for class deepnetts.net.layers.FullyConnectedLayer
-
Creates an instance of fully connected layer with specified number of neurons and ReLU activation function.
- FullyConnectedLayer(int, ActivationType) - Constructor for class deepnetts.net.layers.FullyConnectedLayer
-
Creates an instance of a fully connected layer with specified width (number of neurons) and activation function type.
G
- gaussian(float[], float, float) - Static method in class deepnetts.net.weights.RandomWeights
- GAUSSIAN - Enum constant in enum class deepnetts.net.weights.RandomWeightsType
- generateNoisyImage(int, int, int, String) - Static method in class deepnetts.util.ImageUtils
- generateRandomColoredImages(int, int, int, String) - Static method in class deepnetts.util.ImageUtils
-
Generates a specified number of randomly full colored images of a specified size.
- generateRandomFilledRectImages(int, int, int, String) - Static method in class deepnetts.util.ImageUtils
- GenerateRandomNegative - Class in deepnetts.util
- GenerateRandomNegative() - Constructor for class deepnetts.util.GenerateRandomNegative
- get(int) - Method in class deepnetts.tensor.Tensor1D
-
Gets value from specified index position.
- get(int, int) - Method in class deepnetts.eval.ConfusionMatrix
-
Returns a value of confusion matrix at specified position.
- get(int, int) - Method in class deepnetts.tensor.Tensor2D
-
Returns a value at specified row, col position in this tensor.
- get(int, int, int) - Method in class deepnetts.tensor.Tensor3D
- get(int, int, int, int) - Method in class deepnetts.tensor.Tensor4D
- get(String) - Method in class deepnetts.automl.Parameters
- get(String) - Method in class deepnetts.util.Parameters
- getAcceleratorBridge() - Method in class deepnetts.tensor.TensorBase
- getAccuracy() - Method in class deepnetts.eval.ClassificationMetrics
-
Percent of correct classifications (for both positive and negative classes).
- getActivation() - Method in class deepnetts.net.layers.AbstractLayer
- getActivationType() - Method in class deepnetts.net.layers.AbstractLayer
- getAsProperties() - Method in class deepnetts.automl.Parameters.ParameterCombination
- getBackwardAcc() - Method in class deepnetts.net.layers.AbstractLayer
- getBalancedAccuracy() - Method in class deepnetts.eval.ClassificationMetrics
-
Balanced accuracy is a good metric to use when data set is not balanced.
- getBatchSize() - Method in class deepnetts.net.layers.AbstractLayer
- getBatchSize() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Batch size is number of training examples after which network's weights are adjusted.
- getBestNetwork() - Method in class deepnetts.net.train.KFoldCrossValidation
- getBestResult() - Method in class deepnetts.net.train.KFoldCrossValidation
- getBiases() - Method in class deepnetts.net.layers.AbstractLayer
- getBiasLearningRate() - Method in class deepnetts.net.train.opt.SgdOptimizer
- getBlue(int) - Static method in class deepnetts.util.ColorUtils
- getBoolean(String) - Method in class deepnetts.util.Parameters
- getBoolean(String) - Method in class deepnetts.util.TypedProperties
-
Get specified property as boolean. .
- getBrightness() - Method in class deepnetts.data.ImageSet
- getCheckpointEpochs() - Method in class deepnetts.net.train.BackpropagationTrainer
-
On how many epochs the snapshots of the trained network should be created.
- getClassCount() - Method in class deepnetts.eval.ConfusionMatrix
- getClassIdx() - Method in class deepnetts.eval.ClassificationMetrics
- getClassLabel() - Method in class deepnetts.eval.ClassificationMetrics
-
Returns class label that these metric correspond to (used for multi class classification).
- getClassLabels() - Method in class deepnetts.eval.ConfusionMatrix
- getCol(int, float[]) - Method in class deepnetts.tensor.Tensor2D
- getColorFor(float, float, float) - Static method in class deepnetts.util.ColorUtils
- getColsCache(int) - Method in class deepnetts.tensor.Tensor2D
- getColumnNames() - Method in class deepnetts.data.TabularDataSet
- getColumnNames() - Method in class deepnetts.util.CsvFormat
- getColumnTypes() - Method in class deepnetts.util.CsvFormat
- getConfusionMatrix() - Method in class deepnetts.eval.ClassificationMetrics
-
Returns a confusion matrix that is used to generate these metrics.
- getConfusionMatrix() - Method in class deepnetts.eval.ClassifierEvaluator
- getCropCornersAndCenter() - Method in class deepnetts.data.ImageSet
- getCurrentEpoch() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Returns the current training epoch(iteration) of this trainer.
- getDefault() - Static method in class deepnetts.util.RandomGenerator
-
Returns the default global instance of random generator
- getDelimiter() - Method in class deepnetts.data.ImageSet
- getDelimiter() - Method in class deepnetts.util.CsvFormat
- getDeltaBiases() - Method in class deepnetts.net.layers.AbstractLayer
- getDeltas() - Method in class deepnetts.net.layers.AbstractLayer
- getDeltas() - Method in interface deepnetts.net.layers.Layer
-
Returns layer deltas/errors (as a tensor).
- getDeltaWeights() - Method in class deepnetts.net.ConvolutionalNetwork
-
Returns delta weights for all layers.
- getDeltaWeights() - Method in class deepnetts.net.layers.AbstractLayer
- getDepth() - Method in class deepnetts.net.layers.AbstractLayer
- getDim(int) - Method in class deepnetts.tensor.Shape
- getDimensions() - Method in class deepnetts.tensor.Shape
- getDouble(String) - Method in class deepnetts.util.TypedProperties
-
Get specified property as double .
- getDoubleArray(String) - Method in class deepnetts.util.TypedProperties
- getDropout() - Method in class deepnetts.net.layers.FullyConnectedLayer
- getDropout() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Dropout is a technique to prevent overfitting, which skips adjusting weights for some neurons with given probability.
- getEarlyStopping() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Early stopping stops training if it starts converging slow, and prevents overfitting.
- getEarlyStoppingMinLossChange() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Early stopping stops training if the error/loss start converging to slow.
- getEarlyStoppingPatience() - Method in class deepnetts.net.train.BackpropagationTrainer
-
How many epochs to wait to see if the loss is lowering to slow.
- getEmbeddingDim() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- getEmbeddingDim() - Method in class deepnetts.net.layers.EmbeddingLayer
- getEmbeddingDim() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- getEmbeddingDim() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- getEpochs() - Method in class deepnetts.net.train.TrainingResult
- getError() - Method in class deepnetts.data.DataSetItem
- getError() - Method in class deepnetts.data.ExampleImage
- getError() - Method in interface deepnetts.data.MLDataItem
-
Returns a model output error for this data item.
- getErrorRate() - Method in class deepnetts.eval.ClassificationMetrics
-
A percent of wrong classifications/predictions made.
- getEvaluationMetrics() - Method in class deepnetts.net.train.TrainingResult
- getEvaluator() - Method in class deepnetts.automl.HyperParameterSearch
- getExecutorService() - Method in class deepnetts.util.DeepNettsThreadPool
- getExtendedLogging() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Extended logging includes additional info for debugging the training.
- getF1Score() - Method in class deepnetts.eval.ClassificationMetrics
-
Calculates and returns F1 score - a balance between recall and precision. f1 = 2 * ( (precision*recall) / (precision+recall))
- getFalseDiscoveryRate() - Method in class deepnetts.eval.ClassificationMetrics
-
When its actually no, how often it is classified as yes
- getFalseNegative() - Method in class deepnetts.eval.ConfusionMatrix
-
How many positive items has been (falsely) classified as negative.
- getFalseNegative(int) - Method in class deepnetts.eval.ConfusionMatrix
- getFalseNegativeRate() - Method in class deepnetts.eval.ClassificationMetrics
-
When its actually yes, how often does it predicts no
- getFalsePositive() - Method in class deepnetts.eval.ConfusionMatrix
-
Returns number of false positive classifications.
- getFalsePositive(int) - Method in class deepnetts.eval.ConfusionMatrix
- getFalsePositiveRate() - Method in class deepnetts.eval.ClassificationMetrics
-
When it's actually no, how often does it predict yes?
- getFile() - Method in class deepnetts.data.ExampleImage
- getFileHandler() - Method in class deepnetts.util.DeepNettsLogManager
- getFilterDeltaWeights() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getFilterDepth() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getFilterHeight() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getFilterHeight() - Method in class deepnetts.net.layers.MaxPoolingLayer
- getFilters() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getFilterWidth() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getFilterWidth() - Method in class deepnetts.net.layers.MaxPoolingLayer
- getFlipHorizontal() - Method in class deepnetts.data.ImageSet
- getFloat(String) - Method in class deepnetts.util.Parameters
- getFloat(String) - Method in class deepnetts.util.TypedProperties
-
Get specified property as float.
- getFloatArray(String) - Method in class deepnetts.util.TypedProperties
- getForwardAcc() - Method in class deepnetts.net.layers.AbstractLayer
- getFScore(int) - Method in class deepnetts.eval.ClassificationMetrics
-
Balance between precision and recall.
- getFStat() - Method in class deepnetts.eval.RegressionMetrics
- getGradients() - Method in class deepnetts.net.layers.AbstractLayer
- getGrayscale() - Method in class deepnetts.data.ImageSet
- getGreen(int) - Static method in class deepnetts.util.ColorUtils
- getGroups() - Method in class deepnetts.net.layers.Filter
- getHeight() - Method in class deepnetts.data.ExampleImage
- getHeight() - Method in class deepnetts.net.layers.AbstractLayer
- getHeight() - Method in class deepnetts.net.layers.Filter
- getHeight() - Method in class deepnetts.util.BoundingBox
- getId() - Method in class deepnetts.util.BoundingBox
- getId2word() - Method in class deepnetts.tokenizers.WordTokenizerParameters
- getImageHeight() - Method in class deepnetts.data.ImageSet
- getImageType(File) - Static method in class deepnetts.util.ImageUtils
-
Returns the extension of the given image file.
- getImageWidth() - Method in class deepnetts.data.ImageSet
- getInput() - Method in class deepnetts.data.DataSetItem
- getInput() - Method in class deepnetts.data.ExampleImage
- getInput() - Method in interface deepnetts.data.MLDataItem
-
Returns an input for machine learning model of this item.
- getInputLayer() - Method in class deepnetts.net.NeuralNetwork
-
Returns the input layer of this neural network.
- getInstance() - Static method in class deepnetts.core.DeepNetts
-
Returns a singleton instance of a Deep Netts engine.The instance provides global configuration settings for the Deep Netts runtime.
- getInt(String) - Method in class deepnetts.automl.Parameters
- getInt(String) - Method in class deepnetts.util.TypedProperties
-
Get specified property as int.
- getIntArray(String) - Method in class deepnetts.util.TypedProperties
- getInteger(String) - Method in class deepnetts.util.Parameters
- getInvertImages() - Method in class deepnetts.data.ImageSet
- getIqr() - Method in class deepnetts.util.Stats
- getL1RegSum() - Method in class deepnetts.net.NeuralNetwork
-
Calculates and returns L1 regularization sum of the entire network (all layers included).
- getL1Regularization() - Method in class deepnetts.net.layers.AbstractLayer
- getL1WeightSum() - Method in class deepnetts.net.layers.AbstractLayer
- getL1WeightSum() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getL1WeightSum() - Method in class deepnetts.net.layers.MaxPoolingLayer
- getL2RegSum() - Method in class deepnetts.net.NeuralNetwork
-
Calculates and returns L2 regularization sum of the entire network (all layers included).
- getL2Regularization() - Method in class deepnetts.net.layers.AbstractLayer
- getL2WeightSum() - Method in class deepnetts.net.layers.AbstractLayer
- getL2WeightSum() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getL2WeightSum() - Method in class deepnetts.net.layers.MaxPoolingLayer
- getLabel() - Method in class deepnetts.data.ExampleImage
- getLabel() - Method in class deepnetts.net.NeuralNetwork
-
Returns the label(name) of this neural network
- getLabel() - Method in class deepnetts.util.BoundingBox
- getLabels() - Method in class deepnetts.util.LabelProbabilities
- getLabelsCount() - Method in class deepnetts.data.ImageSet
- getLayerAt(int) - Method in class deepnetts.net.NeuralNetwork
- getLayers() - Method in class deepnetts.net.NeuralNetwork
-
Gets layers of this neural network.
- getLayersGradients() - Method in class deepnetts.net.ConvolutionalNetwork
- getLayersOutputs() - Method in class deepnetts.net.ConvolutionalNetwork
-
Returns outputs of all layers.
- getLearningRate() - Method in class deepnetts.net.layers.AbstractLayer
- getLearningRate() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Learning rate controls the step size as a percent of the error to use for adjusting internal parameters(weights) of the neural network.
- getLearningRate() - Method in class deepnetts.net.train.opt.SgdOptimizer
- getLoss() - Method in class deepnetts.net.train.TrainingResult
- getLossFunction() - Method in class deepnetts.net.NeuralNetwork
-
Returns a loss function of this network, which is used to calculate total network error during the training.
- getLossType() - Method in class deepnetts.net.layers.OutputLayer
- getMacroAverage() - Method in class deepnetts.eval.ClassifierEvaluator
- getMacroAverage() - Method in class deepnetts.net.train.KFoldCrossValidation
- getMatthewsCorrelationCoefficient() - Method in class deepnetts.eval.ClassificationMetrics
-
Calculates and returns the matthews corellation coefficient.
- getMax() - Method in class deepnetts.automl.Range
- getMax() - Method in class deepnetts.util.Stats
- getMaxEpochs() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Returns the setting for maximum number of training epochs(iterations).
- getMaxError() - Method in class deepnetts.eval.RegressionMetrics
- getMaxError() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Returns the setting for the stopping error threshold.
- getMaxInputs() - Method in class deepnetts.data.norm.MaxScaler
- getMaxOutputs() - Method in class deepnetts.data.norm.MaxScaler
- getMaxThreads() - Method in class deepnetts.core.DeepNetts
- getMaxThreadsNum() - Static method in class deepnetts.util.DeepNettsThreadPool
- getMean() - Method in class deepnetts.data.ImageSet
- getMean() - Method in class deepnetts.util.ImagePreprocessing
- getMean() - Method in class deepnetts.util.Stats
- getMeanAbsoluteError() - Method in class deepnetts.eval.RegressionMetrics
- getMeanAbsolutePercentageError() - Method in class deepnetts.eval.RegressionMetrics
- getMeanSquaredError() - Method in class deepnetts.eval.RegressionMetrics
-
Mean squared error is the average value of the sum of squared errors.
- getMedian() - Method in class deepnetts.util.Stats
- getMerges() - Method in class deepnetts.tokenizers.bpe.BPETokenizerParameters
- getMetricsByClass() - Method in class deepnetts.eval.ClassifierEvaluator
- getMin() - Method in class deepnetts.automl.Range
- getMin() - Method in class deepnetts.util.Stats
- getMode() - Method in class deepnetts.net.layers.AbstractLayer
- getMode() - Method in class deepnetts.net.NeuralNetwork
- getMomentum() - Method in class deepnetts.net.layers.AbstractLayer
- getMomentum() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Momentum settings helps to avoid oscillations in weight changes and get more stable and faster training.
- getName() - Method in class deepnetts.automl.Parameter
- getName() - Method in class deepnetts.util.Parameter
- getNetwork() - Method in class deepnetts.automl.HyperParameterSearch
- getNetworkType() - Method in class deepnetts.net.layers.AbstractLayer
- getNeuralNetwork() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Returns a neural network trained by this trainer.
- getNextLayer() - Method in class deepnetts.net.layers.AbstractLayer
- getNormalizer() - Method in class deepnetts.net.NeuralNetwork
-
Returns data normalization method that is applied to network's inputs.
- getNumColumns() - Method in class deepnetts.util.CsvFormat
- getNumInputs() - Method in class deepnetts.data.TabularDataSet
- getNumOutputs() - Method in class deepnetts.data.TabularDataSet
- getNumThreads(int) - Method in class deepnetts.net.layers.AbstractLayer
- getOptimizer() - Method in class deepnetts.net.layers.AbstractLayer
- getOptimizer() - Method in class deepnetts.net.train.BackpropagationTrainer
- getOptimizerType() - Method in class deepnetts.net.layers.AbstractLayer
- getOrCreateAccBridge() - Method in class deepnetts.tensor.TensorBase
- getOriginalImageFromFile() - Method in class deepnetts.data.ExampleImage
- getOutput() - Method in class deepnetts.net.NeuralNetwork
-
Returns network's output.
- getOutput(float[]) - Method in class deepnetts.net.FeedForwardNetwork
-
Deprecated.
- getOutputAsTensor() - Method in class deepnetts.net.NeuralNetwork
- getOutputErrors() - Method in class deepnetts.net.layers.OutputLayer
- getOutputLabel(int) - Method in class deepnetts.net.NeuralNetwork
-
Gets a label of the i-th output this network.
- getOutputLabels() - Method in class deepnetts.net.NeuralNetwork
-
Returns all labels for outputs of this network.
- getOutputLayer() - Method in class deepnetts.net.NeuralNetwork
-
Returns the output layer of this network.
- getOutputs() - Method in class deepnetts.net.layers.AbstractLayer
- getOutputs() - Method in interface deepnetts.net.layers.Layer
-
Returns output of this layer (as a tensor).
- getPadding() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getPadding() - Method in class deepnetts.net.layers.Filter
- getParameters() - Method in class deepnetts.automl.HyperParameterSearch
- getParameters() - Method in class deepnetts.automl.Parameters.ParameterCombination
- getPatternLoss() - Method in class deepnetts.net.loss.BinaryCrossEntropyLoss
- getPatternLoss() - Method in class deepnetts.net.loss.CrossEntropyLoss
- getPatternLoss() - Method in interface deepnetts.net.loss.LossFunction
- getPatternLoss() - Method in class deepnetts.net.loss.MeanSquaredErrorLoss
- getPrecision() - Method in class deepnetts.eval.ClassificationMetrics
-
What percent of those predicted as positive are really positive.
- getPreprocessing() - Method in class deepnetts.net.NeuralNetwork
-
Gets preprocessing that needs to be performed before input is fed to this network.
- getPrevDeltaBiases() - Method in class deepnetts.net.layers.AbstractLayer
- getPrevDeltaWeights() - Method in class deepnetts.net.layers.AbstractLayer
- getPrevlayer() - Method in class deepnetts.net.layers.AbstractLayer
- getPrime(float) - Method in interface deepnetts.net.layers.activation.ActivationFunction
-
Returns the first derivative of activation function for specified output y
- getPrime(float) - Method in class deepnetts.net.layers.activation.LeakyRelu
- getPrime(float) - Method in class deepnetts.net.layers.activation.Linear
- getPrime(float) - Method in class deepnetts.net.layers.activation.Relu
- getPrime(float) - Method in class deepnetts.net.layers.activation.Sigmoid
- getPrime(float) - Method in class deepnetts.net.layers.activation.SoftSign
-
Note that this method accepts x as param not y as others
- getPrime(float) - Method in class deepnetts.net.layers.activation.Tanh
- getProbabilityOf(String) - Method in class deepnetts.util.LabelProbabilities
- getProperties() - Method in class deepnetts.core.DeepNetts
-
Returns all the configuration properties for Deep Netts.
- getProperties() - Method in class deepnetts.net.train.TrainingResult
- getProperty(String) - Method in class deepnetts.core.DeepNetts
-
Returns a specified property of the Deep Netts runtime environment.
- getQ1() - Method in class deepnetts.util.Stats
- getQ3() - Method in class deepnetts.util.Stats
- getR2() - Method in class deepnetts.eval.RegressionMetrics
-
Proportion of variance explained by the model.A value between 0 and 1, where 1 is the best and 0 worst.
- getRandom() - Method in class deepnetts.util.RandomGenerator
-
Returns underlying Java random generator
- getRandomNum() - Method in class deepnetts.automl.Range
- getRange() - Method in class deepnetts.util.Stats
- getRecall() - Method in class deepnetts.eval.ClassificationMetrics
-
Ratio between those classified as positive compared to those that are actually positive.
- getRed(int) - Static method in class deepnetts.util.ColorUtils
- getResidualStandardError() - Method in class deepnetts.eval.RegressionMetrics
- getResizeStrategy() - Method in class deepnetts.data.ImageSet
- getResults() - Method in class deepnetts.automl.HyperParameterSearch
- getRgbVector() - Method in class deepnetts.data.ExampleImage
- getRootMeanSquaredError() - Method in class deepnetts.eval.RegressionMetrics
- getRow(int, float[]) - Method in class deepnetts.tensor.Tensor2D
- getRowsCache(int) - Method in class deepnetts.tensor.Tensor2D
- getScaleImages() - Method in class deepnetts.data.ImageSet
-
Returns flag that indicates wheather images should be scaled to specified dimensions while creating image set.
- getScore() - Method in class deepnetts.util.BoundingBox
- getSearchSpace() - Method in class deepnetts.automl.HyperParameterSearch
- getShuffle() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Returns shuffle flag which determines if training set should be shuffled before each epoch.
- getSingleOutInput() - Method in class deepnetts.net.layers.OutputLayer
- getSnapshotEpochs() - Method in class deepnetts.net.train.BackpropagationTrainer
-
On how many epochs to make training snapshots.
- getSnapshotPath() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Path to use for making snapshots - saving the current state of trained network during the training in order to be able to restore it from a training point if needed.
- getSource() - Method in class deepnetts.net.train.TrainingEvent
-
Gets the source of the event.
- getSpecificity() - Method in class deepnetts.eval.ClassificationMetrics
-
Specificity or true negative rate.
- getSquaredErrorSum() - Method in class deepnetts.eval.RegressionMetrics
- getStd() - Method in class deepnetts.util.Stats
- getStep() - Method in class deepnetts.automl.Range
- getStopAccuracy() - Method in class deepnetts.net.train.BackpropagationTrainer
- getStopError() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Alias for getMaxError().
- getStrategy() - Method in class deepnetts.automl.Range
- getStride() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getStride() - Method in class deepnetts.net.layers.Filter
- getStride() - Method in class deepnetts.net.layers.MaxPoolingLayer
- getString(String) - Method in class deepnetts.util.Parameters
- getString(String) - Method in class deepnetts.util.TypedProperties
-
Get specified property as String .
- getTargetColumnsNames() - Method in class deepnetts.data.ImageSet
-
Returns output/image labels.
- getTargetColumnsNames() - Method in class deepnetts.data.TabularDataSet
- getTargetOutput() - Method in class deepnetts.data.DataSetItem
- getTargetOutput() - Method in class deepnetts.data.ExampleImage
- getTargetOutput() - Method in interface deepnetts.data.MLDataItem
-
Returns target output for machine learning model of this item.
- getTensorDim() - Method in class deepnetts.net.layers.InputLayer
- getTestSet() - Method in class deepnetts.data.TrainTestSplit
- getTestSet() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Test set is used after the training to estimate performance of the trained model and generalization ability with new data.
- getThreadNum() - Method in class deepnetts.util.DeepNettsThreadPool
- getThreadPool() - Method in class deepnetts.net.NeuralNetwork
- getThreshold() - Method in class deepnetts.eval.ClassifierEvaluator
- getTotal() - Method in class deepnetts.eval.ClassificationMetrics
-
Returns total number of classifications.
- getTotal() - Method in class deepnetts.eval.RootMeanSquaredError
- getTotal() - Method in class deepnetts.net.loss.BinaryCrossEntropyLoss
- getTotal() - Method in class deepnetts.net.loss.CrossEntropyLoss
- getTotal() - Method in interface deepnetts.net.loss.LossFunction
-
Returns the total error calculated by this loss function.
- getTotal() - Method in class deepnetts.net.loss.MeanSquaredErrorLoss
- getTotalItems() - Method in class deepnetts.eval.ConfusionMatrix
- getTrainer() - Method in class deepnetts.net.NeuralNetwork
-
Returns a training algorithm of this neural network.
- getTrainer() - Method in interface deepnetts.net.train.TrainerProvider
- getTrainingAccuracy() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Accuracy metric which tells us a percent of correct predictions for training set.
- getTrainingListener() - Method in class deepnetts.net.train.KFoldCrossValidation
- getTrainingLoss() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Total training error/loss at the current epoch.
- getTrainingResults() - Method in class deepnetts.net.train.KFoldCrossValidation
- getTrainingSet() - Method in class deepnetts.automl.HyperParameterSearch
- getTrainingSet() - Method in class deepnetts.data.TrainTestSplit
- getTranslate() - Method in class deepnetts.data.ImageSet
- getTransposed() - Method in class deepnetts.tensor.Tensor2D
- getTrueNegative() - Method in class deepnetts.eval.ConfusionMatrix
- getTrueNegative(int) - Method in class deepnetts.eval.ConfusionMatrix
- getTruePositive() - Method in class deepnetts.eval.ConfusionMatrix
-
Return true positive metric for binary classification.
- getTruePositive(int) - Method in class deepnetts.eval.ConfusionMatrix
-
Returns true positive metric for specified class idx for multiclass classification.
- getType() - Method in class deepnetts.net.train.TrainingEvent
-
Gets the type of the event.
- getValidationAccuracy() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Accuracy metric which tells us a percent of correct predictions for validation set.
- getValidationLoss() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Validation loss is an error calculated using validation set, used to prevent overfitting, and validate architecture and training settings.
- getValue() - Method in class deepnetts.automl.Parameter
-
If it does not hold list but a single value
- getValue() - Method in class deepnetts.util.Parameter
- getValue(float) - Method in interface deepnetts.net.layers.activation.ActivationFunction
-
Returns the value of activation function for specified input x
- getValue(float) - Method in class deepnetts.net.layers.activation.LeakyRelu
- getValue(float) - Method in class deepnetts.net.layers.activation.Linear
- getValue(float) - Method in class deepnetts.net.layers.activation.Relu
- getValue(float) - Method in class deepnetts.net.layers.activation.Sigmoid
- getValue(float) - Method in class deepnetts.net.layers.activation.SoftSign
- getValue(float) - Method in class deepnetts.net.layers.activation.Tanh
- getValues() - Method in class deepnetts.automl.Parameter
-
Returns the list of possible values for this parameter
- getValues() - Method in interface deepnetts.tensor.Tensor
-
Returns an array of values/elements stored in this tensor.
- getValues() - Method in class deepnetts.tensor.TensorBase
- getVar() - Method in class deepnetts.util.Stats
- getVocab() - Method in class deepnetts.tokenizers.bpe.BPETokenizerParameters
- getVocabSize() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- getVocabSize() - Method in class deepnetts.net.layers.EmbeddingLayer
- getVocabSize() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- getVocabSize() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- getWeights() - Method in class deepnetts.net.ConvolutionalNetwork
-
Returns weights from all layers in this network as a list of tensors.
- getWeights() - Method in class deepnetts.net.layers.AbstractLayer
- getWeights() - Method in class deepnetts.net.layers.ConvolutionalLayer
- getWidth() - Method in class deepnetts.data.ExampleImage
- getWidth() - Method in class deepnetts.net.layers.AbstractLayer
- getWidth() - Method in class deepnetts.net.layers.Filter
- getWidth() - Method in class deepnetts.util.BoundingBox
- getWord2id() - Method in class deepnetts.tokenizers.WordTokenizerParameters
- getXMin() - Method in class deepnetts.util.BoundingBox
- getYCenter() - Method in class deepnetts.util.BoundingBox
- getYMax() - Method in class deepnetts.util.BoundingBox
- getYMin() - Method in class deepnetts.util.BoundingBox
- getZeroMeanPixels() - Method in class deepnetts.data.ImageSet
- gradients - Variable in class deepnetts.net.layers.AbstractLayer
-
Gradients of a loss function calculates during a backward pass.
- gradients - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- grayscale(BufferedImage) - Static method in class deepnetts.util.ImageUtils
-
Returns grayscale version of the given image.
- GRID - Enum constant in enum class deepnetts.automl.SearchStrategy
- GRID - Static variable in class deepnetts.automl.HyperParameterSearch
- groups(int) - Method in class deepnetts.net.layers.Filter
H
- HALVING - Enum constant in enum class deepnetts.automl.SearchStrategy
- handleEvent(TrainingEvent) - Method in class deepnetts.net.train.opt.AdaDeltaOptimizer
- handleEvent(TrainingEvent) - Method in class deepnetts.net.train.opt.AdamOptimizer
- handleEvent(TrainingEvent) - Method in class deepnetts.net.train.opt.LearningRateDecay
- handleEvent(TrainingEvent) - Method in interface deepnetts.net.train.TrainingListener
-
Invoked when a training event occurs.
- hashCode() - Method in class deepnetts.automl.Parameter
- hashCode() - Method in class deepnetts.automl.Parameters.ParameterCombination
- hashCode() - Method in class deepnetts.tensor.Shape
- hashCode() - Method in class deepnetts.tensor.TensorBase
- hashCode() - Method in record class deepnetts.tokenizers.bpe.Pair
-
Returns a hash code value for this object.
- hasMissingValues() - Method in class deepnetts.data.TabularDataSet
- hasMissingValues(int) - Method in class deepnetts.data.TabularDataSet
- he(float[], int) - Static method in class deepnetts.net.weights.RandomWeights
-
He initialization, used for relu activations.zero-mean.
- HE - Enum constant in enum class deepnetts.net.weights.RandomWeightsType
- height - Variable in class deepnetts.net.layers.AbstractLayer
- HIDDEN_ACTIVATION - Static variable in class deepnetts.automl.Parameters
- HIDDEN_LAYERS - Static variable in class deepnetts.automl.Parameters
- HIDDEN_NEURONS - Static variable in class deepnetts.automl.Parameters
- hiddenActivationFunction(ActivationType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Sets default type of the activation function to use for all hidden layers in the network.
- hiddenActivationFunction(ActivationType) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Sets default type of the activation function to use for all hidden layers in the network.
- hIdx - Variable in class deepnetts.tensor.Layout
- histogram(float[]) - Static method in class deepnetts.util.Statistics
- histogram(float[], int) - Static method in class deepnetts.util.Statistics
- histogram(float[], int, float, float) - Static method in class deepnetts.util.Statistics
- Histogram() - Constructor for class deepnetts.util.Statistics.Histogram
- HyperParameterSearch - Class in deepnetts.automl
-
The core automl class that performs automated model building and evaluation with specified parameters.
- HyperParameterSearch() - Constructor for class deepnetts.automl.HyperParameterSearch
I
- IMAGE_IDX_FILE - Static variable in class deepnetts.util.ImageSetUtils
- ImagePreprocessing - Class in deepnetts.util
-
Image pre-processing to be used at inference time, after feeding input to the network.
- ImagePreprocessing() - Constructor for class deepnetts.util.ImagePreprocessing
- ImageResize - Enum Class in deepnetts.util
- ImageSet - Class in deepnetts.data
-
Data set with images that will be used to train convolutional neural network.
- ImageSet(int, int) - Constructor for class deepnetts.data.ImageSet
-
Creates empty image set for images of specified size.
- ImageSet(int, int, String) - Constructor for class deepnetts.data.ImageSet
-
Creates image set with images from specified directory path.
- ImageSetUtils - Class in deepnetts.util
- ImageSetUtils() - Constructor for class deepnetts.util.ImageSetUtils
- ImageUtils - Class in deepnetts.util
-
Utility methods to work with images.
- importWeights(NeuralNetwork, String) - Static method in class deepnetts.util.TensorflowUtils
- inc(int, int) - Method in class deepnetts.eval.ConfusionMatrix
-
Increments matrix value at specified position.
- INFERENCE - Enum constant in enum class deepnetts.net.Mode
- init() - Method in class deepnetts.net.layers.AbstractLayer
-
This method should implement layer initialization in subclasses, when a layer is added to the network (create weights, outputs, deltas, randomization etc.).
- init() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- init() - Method in class deepnetts.net.layers.ConvolutionalLayer
-
Initialize dimensions, create output tensors, filters/weights, biases and all internal structures etc.
- init() - Method in class deepnetts.net.layers.EmbeddingLayer
- init() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- init() - Method in class deepnetts.net.layers.FlattenLayer
- init() - Method in class deepnetts.net.layers.FullyConnectedLayer
-
Creates all internal data structures: inputs, weights, biases, outputs, deltas, deltaWeights, deltaBiases prevDeltaWeights, prevDeltaBiases.
- init() - Method in class deepnetts.net.layers.InputLayer
-
Initialize this layer in network.
- init() - Method in class deepnetts.net.layers.LayerNorm
- init() - Method in class deepnetts.net.layers.MaxPoolingLayer
- init() - Method in class deepnetts.net.layers.OutputLayer
- init() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- init() - Method in class deepnetts.net.layers.SoftmaxOutputLayer
- initBuffers(int, int, int) - Method in class deepnetts.net.layers.LayerNorm
- initSeed(long) - Static method in class deepnetts.net.weights.RandomWeights
-
Initializes random number generator with specified seed.
- initSeed(long) - Method in class deepnetts.util.RandomGenerator
-
Initialize the underlying Java random number generator with specified seed.
- initTransientFields() - Method in class deepnetts.net.layers.AbstractLayer
- initTransientFields() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- initTransientFields() - Method in class deepnetts.net.layers.ConvolutionalLayer
- initTransientFields() - Method in class deepnetts.net.layers.EmbeddingLayer
- initTransientFields() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- initTransientFields() - Method in class deepnetts.net.layers.FlattenLayer
- initTransientFields() - Method in class deepnetts.net.layers.FullyConnectedLayer
- initTransientFields() - Method in class deepnetts.net.layers.LayerNorm
- initTransientFields() - Method in class deepnetts.net.layers.MaxPoolingLayer
- initTransientFields() - Method in class deepnetts.net.layers.OutputLayer
- initTransientFields() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- initTransientFields() - Method in class deepnetts.net.layers.SoftmaxOutputLayer
- initWeights(int) - Method in class deepnetts.net.layers.LayerNorm
- INPUT - Enum constant in enum class deepnetts.net.layers.LayerType
-
Accepts external input to the network.
- InputLayer - Class in deepnetts.net.layers
-
Input layer in a neural network.
- InputLayer(int) - Constructor for class deepnetts.net.layers.InputLayer
-
Creates input layer with specified width, and with height and depth equals to one.
- InputLayer(int, int) - Constructor for class deepnetts.net.layers.InputLayer
-
Creates input layer with specified width and height, and depth=1 (single depth/channel).
- InputLayer(int, int, boolean) - Constructor for class deepnetts.net.layers.InputLayer
- InputLayer(int, int, int) - Constructor for class deepnetts.net.layers.InputLayer
-
Creates input layer with specified width, height, and depth (number of channels).
- InputLayer(int, int, int, int) - Constructor for class deepnetts.net.layers.InputLayer
- inputs - Variable in class deepnetts.net.layers.AbstractLayer
-
Inputs to this layer.
- inputs - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- INPUTS - Static variable in class deepnetts.automl.Parameters
- INTEGER - Enum constant in enum class deepnetts.util.ColumnType
- invert() - Method in class deepnetts.data.ExampleImage
- isAllocatedOnDev() - Method in interface deepnetts.accl.AcceleratorTensorBridge
- isBatchMode() - Method in class deepnetts.net.layers.AbstractLayer
- isBatchMode() - Method in class deepnetts.net.train.BackpropagationTrainer
-
In batch mode weights are adjusted after the pass of all examples from the training set, while in online mode weights are adjusted after each training example.
- isDebugMode() - Method in class deepnetts.core.DeepNetts
- isEnabled() - Method in class deepnetts.util.ImagePreprocessing
- isHasHeader() - Method in class deepnetts.util.CsvFormat
- isImageFile(File) - Method in class deepnetts.util.AugmentImagesTest
- isImageFile(File) - Method in class deepnetts.util.DeleteAugmentImages
- isImageFile(File) - Method in class deepnetts.util.RandomlyTranslateImages
- isInvertPixels() - Method in class deepnetts.util.ImagePreprocessing
- isMultithreaded() - Method in class deepnetts.core.DeepNetts
- isSubMean() - Method in class deepnetts.util.ImagePreprocessing
- isTrainable() - Method in class deepnetts.net.layers.AbstractLayer
- ITERATION_FINISHED - Enum constant in enum class deepnetts.net.train.TrainingEvent.Type
-
Specifies a training iteration for (single data item) finished event
- ITERATION_FINISHED - Static variable in class deepnetts.net.train.TrainingEvent
-
Specifies that training iteration has finished.
K
- KFoldCrossValidation - Class in deepnetts.net.train
-
Split data set into k parts of equal sizes (folds), then train model with k-1 folds, and validate with remaining 1 fold.
- KFoldCrossValidation.Builder - Class in deepnetts.net.train
-
Builder object for KFoldCrossValidation.
L
- LabelProbabilities - Class in deepnetts.util
- LabelProbabilities(Map) - Constructor for class deepnetts.util.LabelProbabilities
- labels - Variable in class deepnetts.net.layers.OutputLayer
- LABELS_FILE - Static variable in class deepnetts.util.ImageSetUtils
- labelsFromSubDirectories(String) - Static method in class deepnetts.util.ImageSetUtils
-
Returns a list of category/class labels from the names of subdirectories for the given path.
- layer - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- Layer<O> - Interface in deepnetts.net.layers
-
Common base interface for all types of neural network layers.
- LayerNorm - Class in deepnetts.net.layers
- LayerNorm(int) - Constructor for class deepnetts.net.layers.LayerNorm
- LayerNorm(int, float) - Constructor for class deepnetts.net.layers.LayerNorm
- LayerType - Enum Class in deepnetts.net.layers
-
Supported types of layers.
- Layout - Class in deepnetts.tensor
- Layout(int, int, int, int) - Constructor for class deepnetts.tensor.Layout
- LEAKY_RELU - Enum constant in enum class deepnetts.net.layers.activation.ActivationType
- leakyRelu(Tensor, float) - Method in interface deepnetts.accl.spi.ActivationVectorizationProvider
- LeakyRelu - Class in deepnetts.net.layers.activation
-
Leaky Rectified Linear Activation and its Derivative.
- LeakyRelu() - Constructor for class deepnetts.net.layers.activation.LeakyRelu
- LeakyRelu(float) - Constructor for class deepnetts.net.layers.activation.LeakyRelu
- LEARNING_RATE - Static variable in class deepnetts.automl.Parameters
- learningRate - Variable in class deepnetts.net.layers.AbstractLayer
-
Learning rate for this layer.
- learningRate - Variable in class deepnetts.net.train.opt.AbstractOptimizer
- LearningRateDecay - Class in deepnetts.net.train.opt
-
https://www.coursera.org/learn/deep-neural-network/lecture/hjgIA/learning-rate-decay
- LearningRateDecay(float, float) - Constructor for class deepnetts.net.train.opt.LearningRateDecay
- left() - Method in record class deepnetts.tokenizers.bpe.Pair
-
Returns the value of the
leftrecord component. - Linear - Class in deepnetts.net.layers.activation
-
Linear activation function and its derivative.
- Linear() - Constructor for class deepnetts.net.layers.activation.Linear
- Linear(int) - Constructor for class deepnetts.net.layers.activation.Linear
- LINEAR - Enum constant in enum class deepnetts.net.layers.activation.ActivationType
-
Linear activation is used in output layer for regression tasks, or just passing input forward as it is.
- listOf(Parameter...) - Static method in class deepnetts.util.Parameter
- load(String, Class) - Static method in class deepnetts.net.NeuralNetwork
-
Loads and returns neural network previously saved to a file.
- loadFileImageMapFromDirectory(File) - Static method in class deepnetts.util.ImageUtils
-
Loads JPG, JPEG or PNG images from specified directory and returns them as a map with File object as a key and BufferedImage object as a value.
- loadImages(File) - Method in class deepnetts.data.ImageSet
-
Loads example images with corresponding labels from the specified file.
- loadImages(File, int) - Method in class deepnetts.data.ImageSet
-
Loads specified number of example images with corresponding labels from the specified file.
- loadImages(String) - Method in class deepnetts.data.ImageSet
-
Loads images from the specified image index file.
- loadImagesFromDirectory(File) - Static method in class deepnetts.util.ImageUtils
-
Loads all images from the specified directory, and returns them as a list.
- loadLabels(File) - Method in class deepnetts.data.ImageSet
-
Loads and returns image labels to train neural network from the specified file.These labels will be used to label network's outputs.
- loadLabels(String) - Method in class deepnetts.data.ImageSet
-
Loads and returns image labels to train neural network from the specified file.
- LOGGER - Static variable in class deepnetts.core.DeepNetts
- LOGGER - Static variable in class deepnetts.util.DeepNettsLogManager
- lossFunction(LossType) - Method in class deepnetts.net.CBOWNetwork.Builder
- lossFunction(LossType) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Sets loss function to be used by created neural network.
- lossFunction(LossType) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Sets loss function to be used by created neural network.
- lossFunction(LossType) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- lossFunction(LossType) - Method in class deepnetts.net.SkipGramNetwork.Builder
- LossFunction - Interface in deepnetts.net.loss
-
Base Interface for all loss functions.
- lossType - Variable in class deepnetts.net.layers.OutputLayer
- LossType - Enum Class in deepnetts.net.loss
-
Supported types of Loss Functions in Deep Netts engine.
M
- macroAverage(Collection) - Static method in class deepnetts.eval.ClassifierEvaluator
-
Calculates macro average for the given list of ClassificationMetrics.
- macroAverage(Collection) - Static method in class deepnetts.eval.RegresionEvaluator
- main(String[]) - Static method in class deepnetts.util.AugmentImagesTest
- main(String[]) - Static method in class deepnetts.util.CenterOnWhiteBackground
- main(String[]) - Static method in class deepnetts.util.CreateImageIndex
- main(String[]) - Static method in class deepnetts.util.CreateLabelsIndex
- main(String[]) - Static method in class deepnetts.util.DeepNettsRunner
- main(String[]) - Static method in class deepnetts.util.DeleteAugmentImages
- main(String[]) - Static method in class deepnetts.util.GenerateRandomNegative
- main(String[]) - Static method in class deepnetts.util.ObjectsOnBackgrounds
- main(String[]) - Static method in class deepnetts.util.RandomlyTranslateImages
- main(String[]) - Static method in class deepnetts.util.RunScaleImages
- MathFunctions - Class in deepnetts.net.layers.activation
-
Misc math utility functions.
- matMul(Tensor1D, Tensor1D) - Method in class deepnetts.tensor.Tensor2D
-
Performs matrix multiplication with this 2d tensor and vector , and stores results in result tensor.
- matMul(Tensor2D, Tensor2D) - Method in class deepnetts.tensor.Tensor2D
-
Performs dot product operation on this tensor and matrib, and stores results in result tensor.
- matMulFma(Tensor2D, Tensor1D, Tensor1D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulFma(Tensor2D, Tensor2D, Tensor2D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulFmaParallel(Tensor2D, Tensor1D, Tensor1D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulFmaParallel(Tensor2D, Tensor2D, Tensor2D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulMT(Tensor2D, Tensor2D) - Method in class deepnetts.tensor.Tensor2D
- matMulVectorized(Tensor1D, Tensor1D) - Method in class deepnetts.tensor.Tensor2D
- matMulVectorized(Tensor2D, Tensor2D) - Method in class deepnetts.tensor.Tensor2D
- matMulWithAddVector(Tensor2D, Tensor1D, Tensor1D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulWithAddVector(Tensor2D, Tensor2D, Tensor2D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulWithAddVectorParallel(Tensor2D, Tensor1D, Tensor1D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- matMulWithAddVectorParallel(Tensor2D, Tensor2D, Tensor2D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- max() - Method in class deepnetts.util.Range
- max(float[]) - Static method in class deepnetts.util.Statistics
- max(Tensor, Tensor, Tensor) - Static method in class deepnetts.tensor.Tensors
- MAX_EPOCHS - Static variable in class deepnetts.automl.Parameters
- MAX_ERROR - Static variable in class deepnetts.automl.Parameters
- MAXPOOLING - Enum constant in enum class deepnetts.net.layers.LayerType
-
Downsizes the input.
- MaxPoolingLayer - Class in deepnetts.net.layers
-
This layer performs max pooling operation in convolutional neural network, which scales down output from previous layer by taking max outputs from small predefined filter areas.
- MaxPoolingLayer(int, int, int) - Constructor for class deepnetts.net.layers.MaxPoolingLayer
-
Creates a new max pooling layer with specified filter dimensions and stride.
- MaxPoolingLayer(Filter) - Constructor for class deepnetts.net.layers.MaxPoolingLayer
- MaxScaler - Class in deepnetts.data.norm
-
Performs max normalization, rescales data to corresponding max value in each column.
- MaxScaler(DataSet) - Constructor for class deepnetts.data.norm.MaxScaler
-
Creates a new instance of max normalizer initialized to max values in given data set.
- mean() - Method in class deepnetts.tensor.TensorBase
- mean(float[]) - Static method in class deepnetts.util.Statistics
- MEAN_SQUARED_ERROR - Enum constant in enum class deepnetts.net.loss.LossType
-
Mean Squared Error loss, used for regression tasks, implemented by
MeanSquaredErrorLoss - MeanSquaredErrorLoss - Class in deepnetts.net.loss
-
Mean Squared Error Loss function.
- MeanSquaredErrorLoss(NeuralNetwork) - Constructor for class deepnetts.net.loss.MeanSquaredErrorLoss
-
Creates a new mean squared error loss for the given neural network.
- min() - Method in class deepnetts.util.Range
- min(float[]) - Static method in class deepnetts.util.Statistics
- min(Tensor, Tensor, Tensor) - Static method in class deepnetts.tensor.Tensors
- MINI_BATCH - Enum constant in enum class deepnetts.net.train.TrainingEvent.Type
-
Specifies a mini batch (batch of items) finished event
- MINI_BATCH - Static variable in class deepnetts.net.train.TrainingEvent
-
Specifies mini batch event.
- MinMaxScaler - Class in deepnetts.data.norm
-
Performs Min Max normalization on the given data set.
- MinMaxScaler(DataSet) - Constructor for class deepnetts.data.norm.MinMaxScaler
-
Creates a new instance of max normalizer initialized to max values in given data set.
- missingValuesCount(float[]) - Static method in class deepnetts.util.Statistics
- missingValuesCount(float[][]) - Static method in class deepnetts.util.Statistics
- missingValuesCount(TabularDataSet) - Static method in class deepnetts.util.Statistics
- MLDataItem - Interface in deepnetts.data
-
Single data item that will be used to train machine learning model.
- mode - Variable in class deepnetts.net.layers.AbstractLayer
- Mode - Enum Class in deepnetts.net
- model(NeuralNetwork) - Method in class deepnetts.net.train.KFoldCrossValidation.Builder
- momentum - Variable in class deepnetts.net.layers.AbstractLayer
- MOMENTUM - Enum constant in enum class deepnetts.net.train.opt.OptimizerType
- MOMENTUM - Static variable in class deepnetts.automl.Parameters
- MomentumOptimizer - Class in deepnetts.net.train.opt
-
Momentum optimization adds momentum parameter to basic Stochastic Gradient Descent, which can accelerate the process.
- MomentumOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.MomentumOptimizer
- mserror - Variable in class deepnetts.eval.ClassificationMetrics.Stats
- multiply(float) - Method in class deepnetts.tensor.TensorBase
-
Multiplies all the values in tensor with a specified input parameter.
- multiply(float[], float[]) - Static method in class deepnetts.tensor.Tensors
- multiply(float, int) - Method in class deepnetts.tensor.TensorBase
- multiplyElementWise(TensorBase) - Method in class deepnetts.tensor.TensorBase
- multithreaded - Variable in class deepnetts.net.layers.OutputLayer
N
- names() - Method in class deepnetts.automl.Parameters
- NCHW - Static variable in class deepnetts.tensor.Layout
- negativeFreqency() - Method in class deepnetts.eval.ClassificationMetrics
-
How often does negative class actually occur in the sample
- NETWORK_FILE_EXT - Static variable in class deepnetts.util.FileIO
- networkFactory(NetworkFactory) - Method in class deepnetts.automl.HyperParameterSearch
- NetworkFactory<T> - Interface in deepnetts.automl
-
Base interface for all network factories.
- networkType - Variable in class deepnetts.net.layers.AbstractLayer
- NetworkType - Enum Class in deepnetts.net
-
Neural network architecture types.
- NeuralNetwork<T> - Class in deepnetts.net
-
Base class for all neural networks in Deep Netts.
- NeuralNetwork() - Constructor for class deepnetts.net.NeuralNetwork
- nextFloat() - Method in class deepnetts.util.RandomGenerator
- nextGaussian() - Method in class deepnetts.util.RandomGenerator
- nextInt() - Method in class deepnetts.util.RandomGenerator
- nextLayer - Variable in class deepnetts.net.layers.AbstractLayer
-
Next layer in network.
- NHWC - Static variable in class deepnetts.tensor.Layout
- nIdx - Variable in class deepnetts.tensor.Layout
- normal(float[]) - Static method in class deepnetts.net.weights.RandomWeights
- normalizeInput(TensorBase) - Method in class deepnetts.data.norm.MaxScaler
- numBins() - Method in class deepnetts.util.Statistics.Histogram
- numDimensions - Variable in class deepnetts.tensor.TensorBase
- numDimensions() - Method in class deepnetts.tensor.Shape
- numDimensions() - Method in interface deepnetts.tensor.Tensor
-
Returns a number of dimensions of this tensor
- numDimensions() - Method in class deepnetts.tensor.TensorBase
-
Rank corresponds to number of dimensions in tensor.
- numElements() - Method in interface deepnetts.tensor.Tensor
-
Returns a total number of elements in this tensor.
- numElements() - Method in class deepnetts.tensor.TensorBase
-
Total number of values in tensor.
- numSplits(int) - Method in class deepnetts.net.train.KFoldCrossValidation.Builder
- numThreads - Variable in class deepnetts.net.layers.AbstractLayer
O
- ObjectsOnBackgrounds - Class in deepnetts.util
-
Center images on backgounds and save at target path.
- ObjectsOnBackgrounds() - Constructor for class deepnetts.util.ObjectsOnBackgrounds
- of(float[]) - Static method in class deepnetts.tensor.Tensor1D
-
Create and return a 1D tensor of specified values.
- of(float, float) - Static method in class deepnetts.util.Range
- of(int...) - Static method in class deepnetts.tensor.Shape
- of(Class) - Static method in enum class deepnetts.net.loss.LossType
- of(T, T) - Static method in class deepnetts.automl.Range
- of(T, T, T) - Static method in class deepnetts.automl.Range
- Of(Class) - Static method in enum class deepnetts.net.NetworkType
- ofSize(int) - Static method in class deepnetts.net.layers.Filters
-
Factory method that creates filter settings with specified size (using same size for filter width and height).
- ofSize(int, int) - Static method in class deepnetts.net.layers.Filters
-
Factory method that creates a filter settings with specified width and height
- oneHotEncode(String, String[]) - Static method in class deepnetts.data.DataSets
-
Returns one hot encoded vector for the given label.
- ones(int) - Static method in class deepnetts.tensor.Tensors
- optimizer - Variable in class deepnetts.net.layers.AbstractLayer
- Optimizer - Interface in deepnetts.net.train.opt
-
Optimization technique to tune network's weights parameters used by training algorithm.
- OPTIMIZER - Static variable in class deepnetts.automl.Parameters
- optimizerType - Variable in class deepnetts.net.layers.AbstractLayer
- OptimizerType - Enum Class in deepnetts.net.train.opt
-
Supported types of optimization methods used by back-propagation training algorithm.
- outerProduct(Tensor1D, Tensor2D) - Method in class deepnetts.tensor.Tensor1D
- outerProductAccumulate(Tensor2D, Tensor2D) - Method in class deepnetts.tensor.Tensor2D
- outerProductVectorized(Tensor1D, Tensor1D, Tensor2D) - Method in interface deepnetts.accl.spi.TensorVectorizationProvider
- OUTPUT - Enum constant in enum class deepnetts.net.layers.LayerType
-
Provides final output of the network.
- OUTPUT_ACTIVATION - Static variable in class deepnetts.automl.Parameters
- outputErrors - Variable in class deepnetts.net.layers.OutputLayer
- OutputLayer - Class in deepnetts.net.layers
-
Output layer of a neural network.
- OutputLayer(int) - Constructor for class deepnetts.net.layers.OutputLayer
-
Creates an instance of output layer with specified width (number of outputs) and sigmoid activation function by default.
- OutputLayer(int, ActivationType) - Constructor for class deepnetts.net.layers.OutputLayer
-
Creates an instance of output layer with specified width (number of outputs) and specified activation function.
- OutputLayer(String[], ActivationType) - Constructor for class deepnetts.net.layers.OutputLayer
-
Creates an instance of output layer with specified width (number of outputs) and linear activation function by default.
- outputs - Variable in class deepnetts.net.layers.AbstractLayer
-
Layer outputs.
- OUTPUTS - Static variable in class deepnetts.automl.Parameters
P
- padding(int) - Method in class deepnetts.net.layers.Filter
- Pair<L,
R> - Record Class in deepnetts.tokenizers.bpe - Pair(L, R) - Constructor for record class deepnetts.tokenizers.bpe.Pair
-
Creates an instance of a
Pairrecord class. - Parameter<T> - Class in deepnetts.automl
-
A single parameter with name and all possible values to try.
- Parameter<T> - Class in deepnetts.util
- Parameter(String, List) - Constructor for class deepnetts.automl.Parameter
- Parameter(String, T) - Constructor for class deepnetts.util.Parameter
- ParameterCombination() - Constructor for class deepnetts.automl.Parameters.ParameterCombination
- ParameterCombination(List) - Constructor for class deepnetts.automl.Parameters.ParameterCombination
- Parameters - Class in deepnetts.automl
-
Parameter search space: a collection of parameters and methods for generating all possible combinations.
- Parameters - Class in deepnetts.util
-
TODO: setMethod, throw meannigfull exceptions
- Parameters() - Constructor for class deepnetts.automl.Parameters
- Parameters() - Constructor for class deepnetts.util.Parameters
- Parameters.ParameterCombination - Class in deepnetts.automl
-
A single combination of parameters with list of parameters (with all possible values).
- paramRange(String, Range) - Method in class deepnetts.automl.HyperParameterSearch
- paramValue(String, T) - Method in class deepnetts.automl.HyperParameterSearch
-
Set searvh param with specific value.
- paramValues(String, List) - Method in class deepnetts.automl.HyperParameterSearch
- parseFloats(String) - Static method in class deepnetts.tensor.Tensors
- parseInts(String) - Static method in class deepnetts.tensor.Tensors
- physicalCoreCount() - Static method in class deepnetts.util.PhysicalCores
-
Returns the number of "physical" hardware threads available.
- PhysicalCores - Class in deepnetts.util
-
Static utility class for finding the number of physical CPU cores.
- positiveFreqency() - Method in class deepnetts.eval.ClassificationMetrics
-
How often does positive class actually occur in the sample
- precision - Variable in class deepnetts.eval.ClassificationMetrics.Stats
- predict(float...) - Method in class deepnetts.net.FeedForwardNetwork
-
Returns the network's prediction (outputs) for the given input.
- predict(TensorBase) - Method in class deepnetts.net.NeuralNetwork
-
Returns the prediction of this neural network for the given input.
- Preprocessing<T> - Interface in deepnetts.data
-
Data pre-processing abstraction.
- prevDeltaBiases - Variable in class deepnetts.net.layers.AbstractLayer
- prevDeltaWeights - Variable in class deepnetts.net.layers.AbstractLayer
-
Weight changes for current and previous iteration.
- prevLayer - Variable in class deepnetts.net.layers.AbstractLayer
-
Previous layer in network.
- printOutputs(NeuralNetwork) - Static method in class deepnetts.util.Debugger
- printWeights(NeuralNetwork) - Static method in class deepnetts.util.Debugger
- PROP_BATCH_MODE - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the batchMode property
- PROP_BATCH_SIZE - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the batchSize property
- PROP_IMAGE_HEIGHT - Static variable in class deepnetts.util.DeepNettsRunner
- PROP_IMAGE_WIDTH - Static variable in class deepnetts.util.DeepNettsRunner
- PROP_LABELS_FILE - Static variable in class deepnetts.util.DeepNettsRunner
- PROP_LEARNING_RATE - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the learningRate property
- PROP_LEARNING_RATE - Static variable in class deepnetts.util.DeepNettsRunner
- PROP_MAX_EPOCHS - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the maxEpochs property
- PROP_MAX_ERROR - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the maxError property
- PROP_MAX_ERROR - Static variable in class deepnetts.util.DeepNettsRunner
- PROP_MOMENTUM - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the momentum property
- PROP_OPTIMIZER_TYPE - Static variable in class deepnetts.net.train.BackpropagationTrainer
-
Name of the optimizer property
- PROP_TEST_FILE - Static variable in class deepnetts.util.DeepNettsRunner
- PROP_TRAINING_FILE - Static variable in class deepnetts.util.DeepNettsRunner
- provider(String) - Static method in class deepnetts.accl.spi.AcceleratorService
- provider(String) - Static method in class deepnetts.accl.spi.ActivationVectorizationService
- provider(String) - Static method in class deepnetts.accl.spi.TensorVectorizationService
- providers() - Static method in class deepnetts.accl.spi.AcceleratorService
- providers() - Static method in class deepnetts.accl.spi.ActivationVectorizationService
- providers() - Static method in class deepnetts.accl.spi.TensorVectorizationService
- put(Parameter) - Method in class deepnetts.util.Parameters
R
- random(int) - Static method in class deepnetts.tensor.Tensors
-
Generates a random 1D tensor with the specified dimensions.
- random(int, int) - Static method in class deepnetts.tensor.Tensors
-
Create and return a tensor with specified number of rows and cols filled with random values.
- random(int, int, int) - Static method in class deepnetts.tensor.Tensors
- random(int, int, int, int) - Static method in class deepnetts.tensor.Tensors
- RANDOM - Enum constant in enum class deepnetts.automl.SearchStrategy
- RANDOM - Static variable in class deepnetts.automl.HyperParameterSearch
- randomCrop(BufferedImage, int, int) - Static method in class deepnetts.util.ImageUtils
-
Randomly crop image with specified width and height.
- randomCrop(BufferedImage, int, int, int, Random) - Static method in class deepnetts.util.ImageUtils
-
Crops specified number of random sub-images with specified dimensions.
- RandomGenerator - Class in deepnetts.util
-
Random number generator singleton.
- randomize() - Method in class deepnetts.tensor.TensorBase
-
Randomize all values in tensor
- randomize(float[]) - Static method in class deepnetts.net.weights.RandomWeights
-
Initialize the elements of specified array with random numbers with uniform distribution in range [-0.5, 0.5].
- RandomlyTranslateImages - Class in deepnetts.util
-
just move 2(x) pix to left right up down
- RandomlyTranslateImages() - Constructor for class deepnetts.util.RandomlyTranslateImages
- randomNum(int) - Method in class deepnetts.automl.Range
- randomSeed(long) - Method in class deepnetts.automl.HyperParameterSearch
- randomSeed(long) - Method in class deepnetts.net.CBOWNetwork.Builder
- randomSeed(long) - Method in class deepnetts.net.ConvolutionalNetwork.Builder
-
Initializes random number generator with the specified seed in order to get same random number sequences used for weights initialization.
- randomSeed(long) - Method in class deepnetts.net.FeedForwardNetwork.Builder
-
Initializes random number generator with the specified seed in order to get same random number sequences used for weights initialization.
- randomSeed(long) - Method in class deepnetts.net.FeedForwardNetworkLM.Builder
- randomSeed(long) - Method in class deepnetts.net.SkipGramNetwork.Builder
- randomTintAndBrightness(BufferedImage) - Static method in class deepnetts.util.ImageUtils
- randomTintAndBrightness(BufferedImage, float, int, int, Random) - Static method in class deepnetts.util.ImageUtils
-
Still not working as it should
- randomTranslateImage(BufferedImage, int, int) - Static method in class deepnetts.util.ImageUtils
-
Returns an array of images created by translating specified input image by specified number of count with specified step size.
- RandomWeights - Class in deepnetts.net.weights
-
Weights randomization utility methods.
- RandomWeights() - Constructor for class deepnetts.net.weights.RandomWeights
- randomWeightsType - Variable in class deepnetts.net.layers.AbstractLayer
- RandomWeightsType - Enum Class in deepnetts.net.weights
-
Types of supported weights randomization techniques.
- range(float[]) - Static method in class deepnetts.util.Statistics
- Range<T> - Class in deepnetts.automl
-
A value range of type T for the parameter.
- Range - Class in deepnetts.util
- Range(float, float) - Constructor for class deepnetts.util.Range
- Range(T, T) - Constructor for class deepnetts.automl.Range
- Range(T, T, T) - Constructor for class deepnetts.automl.Range
- RangeConsumer - Interface in deepnetts.util
- RangeScaler - Class in deepnetts.data.norm
-
Normalize data set to specified range.
- RangeScaler(float, float) - Constructor for class deepnetts.data.norm.RangeScaler
-
Creates a new instance of range normalizer initialized to given min and max values.
- readCsv(File, int, int, boolean, String) - Static method in class deepnetts.data.DataSets
-
Creates and returns data set from specified CSV file.
- readCsv(String, int, int) - Static method in class deepnetts.data.DataSets
-
Create data set from CSV file, using coma (,) as default delimiter and no header (column names) in first row.
- readCsv(String, int, int, boolean) - Static method in class deepnetts.data.DataSets
- readCsv(String, int, int, boolean, String) - Static method in class deepnetts.data.DataSets
- readCsv(String, int, int, String) - Static method in class deepnetts.data.DataSets
- recall - Variable in class deepnetts.eval.ClassificationMetrics.Stats
- regL1 - Variable in class deepnetts.net.layers.AbstractLayer
- regL2 - Variable in class deepnetts.net.layers.AbstractLayer
- RegresionEvaluator - Class in deepnetts.eval
-
Evaluates regressor neural network for specified data set.
- RegresionEvaluator() - Constructor for class deepnetts.eval.RegresionEvaluator
- RegressionMetrics - Class in deepnetts.eval
-
Common metrics for regression models.
- RegressionMetrics() - Constructor for class deepnetts.eval.RegressionMetrics
- relu(Tensor) - Method in interface deepnetts.accl.spi.ActivationVectorizationProvider
- Relu - Class in deepnetts.net.layers.activation
-
Rectified Linear Activation and its Derivative.
- Relu() - Constructor for class deepnetts.net.layers.activation.Relu
- RELU - Enum constant in enum class deepnetts.net.layers.activation.ActivationType
- removeAllListeners() - Method in class deepnetts.net.train.BackpropagationTrainer
- removeHandler(Handler) - Method in class deepnetts.util.DeepNettsLogManager
- removeListener(TrainingListener) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Removes training listener from this trainer.
- renameFilesAsClasses(String, String) - Static method in class deepnetts.util.ImageSetUtils
-
Renames files in specified directory.
- replace(float, float) - Method in class deepnetts.tensor.TensorBase
- reset() - Method in class deepnetts.net.loss.BinaryCrossEntropyLoss
- reset() - Method in class deepnetts.net.loss.CrossEntropyLoss
- reset() - Method in interface deepnetts.net.loss.LossFunction
-
Resets the total error and pattern counter.
- reset() - Method in class deepnetts.net.loss.MeanSquaredErrorLoss
- ResidualLayer - Class in deepnetts.net.layers
- ResidualLayer() - Constructor for class deepnetts.net.layers.ResidualLayer
- rgbTensor - Variable in class deepnetts.data.ExampleImage
- right() - Method in record class deepnetts.tokenizers.bpe.Pair
-
Returns the value of the
rightrecord component. - RMSPROP - Enum constant in enum class deepnetts.net.train.opt.OptimizerType
- RmsPropOptimizer - Class in deepnetts.net.train.opt
-
A variation of AdaDelta optimizer.
- RmsPropOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.RmsPropOptimizer
- RootMeanSquaredError - Class in deepnetts.eval
-
A measure of error for regression tasks.
- RootMeanSquaredError() - Constructor for class deepnetts.eval.RootMeanSquaredError
- ROW_IDX - Static variable in interface deepnetts.net.train.opt.Optimizer
- ROW_IDX - Static variable in class deepnetts.tensor.Shape
- rows() - Method in class deepnetts.tensor.Tensor2D
-
Returns a number of rows in this matrix.
- rows() - Method in class deepnetts.tensor.Tensor3D
- rows() - Method in class deepnetts.tensor.Tensor4D
- run() - Method in class deepnetts.automl.HyperParameterSearch
- run() - Method in class deepnetts.net.train.KFoldCrossValidation
- run() - Method in class deepnetts.util.AugmentImagesTest
- run() - Method in class deepnetts.util.DeleteAugmentImages
- run() - Method in class deepnetts.util.RandomlyTranslateImages
- run() - Method in class deepnetts.util.RunScaleImages
- run(Runnable) - Method in class deepnetts.util.DeepNettsThreadPool
- run(String) - Method in class deepnetts.util.DeepNettsRunner
- run(Collection) - Method in class deepnetts.util.DeepNettsThreadPool
- RunScaleImages - Class in deepnetts.util
- RunScaleImages() - Constructor for class deepnetts.util.RunScaleImages
S
- save(String) - Method in class deepnetts.net.NeuralNetwork
-
Saves this network using serialization to file with specified fileName.
- scaleAndCenter(BufferedImage, int, int, int, Color) - Static method in class deepnetts.util.ImageUtils
-
Scales input image to specified target width or height, centers and returns resulting image.
- scaleBySmallerAndCrop(BufferedImage, int, int) - Static method in class deepnetts.util.ImageUtils
-
Scales input image to specified target width or height, crops and returns resulting image.
- scaleBySmallerTarget(BufferedImage, int, int) - Static method in class deepnetts.util.ImageUtils
- scaleImage(BufferedImage, int, int) - Static method in class deepnetts.util.ImageUtils
-
Scales specified image to given size and returns new image with specified width and height.
- scaleInput(TensorBase) - Method in class deepnetts.data.norm.AbstractScaler
-
Normalize input of deployed model
- scaleInput(TensorBase) - Method in class deepnetts.data.norm.DecimalScaler
- scaleInput(TensorBase) - Method in class deepnetts.data.norm.MaxScaler
- scaleInput(TensorBase) - Method in class deepnetts.data.norm.MinMaxScaler
- scaleInput(TensorBase) - Method in class deepnetts.data.norm.RangeScaler
- scaleInput(TensorBase) - Method in class deepnetts.data.norm.Standardizer
- scaleToMax(DataSet) - Static method in class deepnetts.data.DataSets
- scaleToMinMax(DataSet) - Static method in class deepnetts.data.DataSets
- SearchStrategy - Enum Class in deepnetts.automl
-
Strategy for hyper-parameter search.
- set(float, int) - Method in class deepnetts.tensor.Tensor1D
-
Sets value at specified index position.
- set(float, int, int) - Method in class deepnetts.tensor.Tensor2D
-
Set value at specified [row, col] position in this tensor.
- set(float, int, int, int) - Method in class deepnetts.tensor.Tensor3D
- set(float, int, int, int, int) - Method in class deepnetts.tensor.Tensor4D
- setActivationType(ActivationType) - Method in class deepnetts.net.layers.AbstractLayer
- setBatchMode(boolean) - Method in class deepnetts.net.layers.AbstractLayer
- setBatchMode(boolean) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Sets flag whether to use batch mode during the training.
- setBatchSize(int) - Method in class deepnetts.net.layers.AbstractLayer
- setBatchSize(int) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Batch size is number of training examples after which network's weights are adjusted.
- setBiases(Tensor1D) - Method in class deepnetts.net.layers.AbstractLayer
- setBiasLearningRate(float) - Method in class deepnetts.net.train.opt.SgdOptimizer
- setBins(int[]) - Method in class deepnetts.util.Statistics.Histogram
- setBinWidth(float) - Method in class deepnetts.util.Statistics.Histogram
- setBrightness(boolean) - Method in class deepnetts.data.ImageSet
- setCheckpointEpochs(int) - Method in class deepnetts.net.train.BackpropagationTrainer
-
On how many epochs the snapshots of the trained network should be created.
- setClassLabel(String) - Method in class deepnetts.eval.ClassificationMetrics
-
Sets class label to which this metrics corresponds too
- setCol(int, float[]) - Method in class deepnetts.tensor.Tensor2D
- setColumnNames(String...) - Method in class deepnetts.data.TabularDataSet
- setColumnNames(String[]) - Method in class deepnetts.util.CsvFormat
- setColumnTypes(ColumnType[]) - Method in class deepnetts.util.CsvFormat
- setCount(int) - Method in class deepnetts.util.Stats
- setCropCornersAndCenter(boolean) - Method in class deepnetts.data.ImageSet
- setCudaHandles(AcceleratorHandle) - Method in class deepnetts.net.layers.AbstractLayer
- setDebugMode(boolean) - Method in class deepnetts.core.DeepNetts
- setDelimiter(String) - Method in class deepnetts.data.ImageSet
- setDelimiter(String) - Method in class deepnetts.util.CsvFormat
- setDeltas(O) - Method in class deepnetts.net.layers.AbstractLayer
- setDropout(float) - Method in class deepnetts.net.layers.FullyConnectedLayer
- setDropout(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Dropout is a technique to prevent overfitting, which skips adjusting weights for some neurons with given probability.
- setEarlyStopping(boolean) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Early stopping stops training if it starts converging slow, and prevents overfitting.
- setEarlyStoppingMinLossChange(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Early stopping stops training if the error/loss start converging to slow.
- setEarlyStoppingPatience(int) - Method in class deepnetts.net.train.BackpropagationTrainer
-
How many epochs to wait to see if the loss is lowering to slow.
- setEmbeddingDim(int) - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- setEmbeddingDim(int) - Method in class deepnetts.net.layers.EmbeddingLayer
- setEmbeddingDim(int) - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- setEmbeddingDim(int) - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- setEnabled(boolean) - Method in class deepnetts.util.ImagePreprocessing
- setError(float) - Method in class deepnetts.data.DataSetItem
- setError(float) - Method in class deepnetts.data.ExampleImage
- setError(float) - Method in interface deepnetts.data.MLDataItem
- setExtendedLogging(boolean) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Extended logging includes additional info for debugging the training.
- setFileHandler(String) - Method in class deepnetts.util.DeepNettsLogManager
- setFilters(Tensor4D) - Method in class deepnetts.net.layers.ConvolutionalLayer
- setFilters(String) - Method in class deepnetts.net.layers.ConvolutionalLayer
- setFlipHorizontal(boolean) - Method in class deepnetts.data.ImageSet
- setFStat(float) - Method in class deepnetts.eval.RegressionMetrics
- setGrayscale(boolean) - Method in class deepnetts.data.ImageSet
- setHasHeader(boolean) - Method in class deepnetts.util.CsvFormat
- setHiddenActivation(ActivationType) - Method in class deepnetts.automl.FeedForwardNetworkFactory
- setId(int) - Method in class deepnetts.util.BoundingBox
- setInput(float...) - Method in class deepnetts.net.FeedForwardNetwork
-
Sets network's input using given inputs and invokes the calculation of the network for the given input (forward pass).
- setInput(float[]) - Method in class deepnetts.net.NeuralNetwork
-
Deprecated.
- setInput(TensorBase) - Method in class deepnetts.net.ConvolutionalNetwork
- setInput(TensorBase) - Method in class deepnetts.net.layers.InputLayer
-
Sets network input
- setInput(TensorBase) - Method in class deepnetts.net.NeuralNetwork
-
Sets network input and calculates entire network (triggers forward pass).
- setInput(String[]) - Method in class deepnetts.net.FeedForwardNetworkLM
- setInputLayer(InputLayer) - Method in class deepnetts.net.NeuralNetwork
-
Sets input layer of this network during the network building.
- setInvertImages(boolean) - Method in class deepnetts.data.ImageSet
- setInvertPixels(boolean) - Method in class deepnetts.util.ImagePreprocessing
- setL1Regularization(float) - Method in class deepnetts.net.layers.AbstractLayer
- setL1Regularization(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
L1 regularization (sum of abs values) is used to prevent overfitting and too large weights.
- setL2Regularization(float) - Method in class deepnetts.net.layers.AbstractLayer
- setL2Regularization(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
L2 regularization (sum of squares) is used to prevent overfitting and too large weights.
- setLabel(String) - Method in class deepnetts.net.NeuralNetwork
-
Sets label(name) for this neural network.
- setLabel(String) - Method in class deepnetts.util.BoundingBox
- setLearningRate(float) - Method in class deepnetts.net.layers.AbstractLayer
- setLearningRate(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Learning rate controls the step size as a percent of the error to use for adjusting internal parameters(weights) of the neural network.
- setLearningRate(float) - Method in class deepnetts.net.train.opt.AdaDeltaOptimizer
- setLearningRate(float) - Method in class deepnetts.net.train.opt.AdaGradOptimizer
- setLearningRate(float) - Method in class deepnetts.net.train.opt.AdamOptimizer
- setLearningRate(float) - Method in class deepnetts.net.train.opt.MomentumOptimizer
- setLearningRate(float) - Method in interface deepnetts.net.train.opt.Optimizer
- setLearningRate(float) - Method in class deepnetts.net.train.opt.RmsPropOptimizer
- setLearningRate(float) - Method in class deepnetts.net.train.opt.SgdOptimizer
- setLearningRateDecay(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Learning rate decay lowers the learning rate with each epoch by devayRate factor, which may improve error lowering the error.
- setLossFunction(LossFunction) - Method in class deepnetts.net.NeuralNetwork
-
Sets a loss function of this network, which is used to calculate total network error during the training.
- setLossType(LossType) - Method in class deepnetts.automl.FeedForwardNetworkFactory
- setLossType(LossType) - Method in class deepnetts.net.layers.OutputLayer
- setMax(float) - Method in class deepnetts.util.Stats
- setMaxEpochs(long) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Deprecated.Use setStopEpochs instead
- setMaxError(float) - Method in class deepnetts.eval.RegressionMetrics
- setMaxError(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Deprecated.Use setStopError instead
- setMaxInputs(TensorBase) - Method in class deepnetts.data.norm.MaxScaler
- setMaxOutputs(TensorBase) - Method in class deepnetts.data.norm.MaxScaler
- setMaxThreads(int) - Method in class deepnetts.core.DeepNetts
- setMean(float) - Method in class deepnetts.util.Stats
- setMean(TensorBase) - Method in class deepnetts.util.ImagePreprocessing
- setMeanAbsoluteError(float) - Method in class deepnetts.eval.RegressionMetrics
- setMeanAbsolutePercentageError(float) - Method in class deepnetts.eval.RegressionMetrics
- setMeanSquaredError(float) - Method in class deepnetts.eval.RegressionMetrics
- setMedian(float) - Method in class deepnetts.util.Stats
- setMerges(Map) - Method in class deepnetts.tokenizers.bpe.BPETokenizerParameters
- setMin(float) - Method in class deepnetts.util.Stats
- setMode(Mode) - Method in class deepnetts.net.layers.AbstractLayer
- setMode(Mode) - Method in class deepnetts.net.NeuralNetwork
- setMomentum(float) - Method in class deepnetts.net.layers.AbstractLayer
- setMomentum(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Momentum settings helps to avoid oscillations in weight changes and get more stable and faster training.
- setNetworkType(NetworkType) - Method in class deepnetts.net.layers.AbstractLayer
- setNextlayer(AbstractLayer) - Method in class deepnetts.net.layers.AbstractLayer
- setNormalizer(AbstractScaler) - Method in class deepnetts.net.NeuralNetwork
-
Sets normalization data normalization method that is applied to network's inputs.
- setNumBins(int) - Method in class deepnetts.util.Statistics.Histogram
- setNumInputs(int) - Method in class deepnetts.automl.FeedForwardNetworkFactory
- setNumOutputs(int) - Method in class deepnetts.automl.FeedForwardNetworkFactory
- setOptimizer(OptimizerType) - Method in class deepnetts.net.train.BackpropagationTrainer
- setOptimizerType(OptimizerType) - Method in class deepnetts.net.layers.AbstractLayer
- setOptimizerType(OptimizerType) - Method in class deepnetts.net.layers.FlattenLayer
- setOutputActivation(ActivationType) - Method in class deepnetts.automl.FeedForwardNetworkFactory
- setOutputError(TensorBase) - Method in class deepnetts.net.NeuralNetwork
-
Sets the network's output errors, which are a difference between actual(predicted) and target output.
- setOutputErrors(TensorBase) - Method in class deepnetts.net.layers.OutputLayer
- setOutputLabels(String...) - Method in class deepnetts.net.NeuralNetwork
-
Sets output labels of this network.
- setOutputLayer(OutputLayer) - Method in class deepnetts.net.NeuralNetwork
-
Sets output layer of this network during the network building.
- setOutputs(O) - Method in class deepnetts.net.layers.AbstractLayer
- setPreprocessing(Preprocessing) - Method in class deepnetts.net.NeuralNetwork
-
Sets preprocessing that needs to be performed before input is fed to this network.
- setPrevDeltaWeights(W) - Method in class deepnetts.net.layers.AbstractLayer
- setPrevLayer(AbstractLayer) - Method in class deepnetts.net.layers.AbstractLayer
- setProperties(Properties) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Sets properties from available keys in specified prop object.
- setQ1(float) - Method in class deepnetts.util.Stats
- setQ3(float) - Method in class deepnetts.util.Stats
- setR2(float) - Method in class deepnetts.eval.RegressionMetrics
- setResidualStandardError(float) - Method in class deepnetts.eval.RegressionMetrics
- setResizeStrategy(ImageResize) - Method in class deepnetts.data.ImageSet
- setRootMeanSquaredError(float) - Method in class deepnetts.eval.RegressionMetrics
- setRow(float[], int, int) - Method in class deepnetts.tensor.Tensor3D
- setRow(int, float[]) - Method in class deepnetts.tensor.Tensor2D
- setScaleImages(boolean) - Method in class deepnetts.data.ImageSet
- setScore(float) - Method in class deepnetts.util.BoundingBox
- setShape(Shape) - Method in class deepnetts.tensor.TensorBase
- setShuffle(boolean) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Sets shuffle flag which determines if training set should be shuffled before each epoch.
- setSnapshotEpochs(int) - Method in class deepnetts.net.train.BackpropagationTrainer
-
On how many epochs to make training snapshots.
- setSnapshotPath(String) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Path to use for making snapshots - saving the current state of trained network during the training in order to be able to restore it from a training point.
- setSquaredErrorSum(float) - Method in class deepnetts.eval.RegressionMetrics
- setStd(float) - Method in class deepnetts.util.Stats
- setStopAccuracy(float) - Method in class deepnetts.net.train.BackpropagationTrainer
- setStopEpochs(long) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Sets number of epochs/iterations to run the training.
- setStopError(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
The training stops when/if training error reach this value.
- setSubMean(boolean) - Method in class deepnetts.util.ImagePreprocessing
- setTargetOutput(TensorBase) - Method in class deepnetts.data.ExampleImage
- setTestSet(DataSet) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Test set is used after the training to estimate performance of the trained model and generalization ability with new data.
- setThreadNum(int) - Method in class deepnetts.util.DeepNettsThreadPool
- setThreadPool(DeepNettsThreadPool) - Method in class deepnetts.net.layers.AbstractLayer
- setThreadPool(DeepNettsThreadPool) - Method in class deepnetts.tensor.Tensor2D
- setThreshold(float) - Method in class deepnetts.eval.ClassifierEvaluator
- setTrainable(boolean) - Method in class deepnetts.net.layers.AbstractLayer
-
Set trainable to false to freeze learned weights.
- setTrainer(T) - Method in class deepnetts.net.NeuralNetwork
-
Sets the training algorithm of this neural network.
- setTrainer(T) - Method in interface deepnetts.net.train.TrainerProvider
- setTrainingSnapshots(boolean) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Training snapshots save the current state of the trained neural network during the training in order to be able to restore it from a training point if needed.
- setTranslate(boolean) - Method in class deepnetts.data.ImageSet
- setUseCuda(boolean) - Method in class deepnetts.core.DeepNetts
- setUseVectorAPI(boolean) - Method in class deepnetts.core.DeepNetts
- setValues(float...) - Method in class deepnetts.tensor.TensorBase
- setValuesAt(int, float[]) - Method in class deepnetts.tensor.Tensor4D
- setValuesFromString(String) - Method in class deepnetts.tensor.TensorBase
-
Sets tensor values from CSV string.
- setValuesFromStringTransposed(String) - Method in class deepnetts.tensor.Tensor2D
- setValuesFromStringTransposed(String) - Method in class deepnetts.tensor.Tensor4D
- setValuesFromStringTransposed(String) - Method in class deepnetts.tensor.TensorBase
- setVar(float) - Method in class deepnetts.util.Stats
- setVocab(Map) - Method in class deepnetts.tokenizers.bpe.BPETokenizerParameters
- setVocabSize(int) - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- setVocabSize(int) - Method in class deepnetts.net.layers.EmbeddingLayer
- setVocabSize(int) - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- setVocabSize(int) - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- setWeights(Tensor4D) - Method in class deepnetts.net.layers.ConvolutionalLayer
- setWeights(String) - Method in class deepnetts.net.layers.AbstractLayer
- setWeights(List) - Method in class deepnetts.net.ConvolutionalNetwork
-
Sets network's weights for all layers.
- setWeights(W) - Method in class deepnetts.net.layers.AbstractLayer
- setXMax(int) - Method in class deepnetts.util.BoundingBox
- setXMin(int) - Method in class deepnetts.util.BoundingBox
- setYCenter(int) - Method in class deepnetts.util.BoundingBox
- setYMax(int) - Method in class deepnetts.util.BoundingBox
- setYMin(int) - Method in class deepnetts.util.BoundingBox
- SGD - Enum constant in enum class deepnetts.net.train.opt.OptimizerType
-
Stochastic Gradient Descent, a basic type of neural network optimization algorithm.
- SgdOptimizer - Class in deepnetts.net.train.opt
-
Basic Stochastic Gradient Descent optimization algorithm, which iteratively changes weights in order to find minimum of loss function.
- SgdOptimizer(AbstractLayer) - Constructor for class deepnetts.net.train.opt.SgdOptimizer
- shape - Variable in class deepnetts.tensor.TensorBase
- shape() - Method in interface deepnetts.tensor.Tensor
-
Returns a shape of this tensor.
- shape() - Method in class deepnetts.tensor.TensorBase
- Shape - Class in deepnetts.tensor
-
Immutable class that represents Tensor shape.
- Shape(int...) - Constructor for class deepnetts.tensor.Shape
- shuffle() - Method in class deepnetts.data.TabularDataSet
-
Shuffles the data set items using the default random generator.
- shuffle(int) - Method in class deepnetts.data.TabularDataSet
-
Shuffles data set items using java random generator initializes with specified seed
- shutdown() - Method in class deepnetts.util.DeepNettsThreadPool
- shutdownNow() - Method in class deepnetts.util.DeepNettsThreadPool
- sigmoid(Tensor) - Method in interface deepnetts.accl.spi.ActivationVectorizationProvider
- Sigmoid - Class in deepnetts.net.layers.activation
-
Sigmoid activation function
- Sigmoid() - Constructor for class deepnetts.net.layers.activation.Sigmoid
- SIGMOID - Enum constant in enum class deepnetts.net.layers.activation.ActivationType
-
Sigmoid activation is used in output layer for binary classification tasks and logistic regression.
- SKIPGRAM - Enum constant in enum class deepnetts.net.NetworkType
- SkipGramEmbeddingLayer - Class in deepnetts.net.layers
- SkipGramEmbeddingLayer(int, int) - Constructor for class deepnetts.net.layers.SkipGramEmbeddingLayer
- SkipGramNetwork - Class in deepnetts.net
- SkipGramNetwork.Builder - Class in deepnetts.net
- SOFTMAX - Enum constant in enum class deepnetts.net.layers.activation.ActivationType
-
SoftMax activation is typically used in output layer, for multi class classification tasks.
- SoftmaxOutputLayer - Class in deepnetts.net.layers
-
Output layer with softmax activation function.
- SoftmaxOutputLayer(int) - Constructor for class deepnetts.net.layers.SoftmaxOutputLayer
- SoftSign - Class in deepnetts.net.layers.activation
-
Soft sign activation function
- SoftSign() - Constructor for class deepnetts.net.layers.activation.SoftSign
- split(double...) - Method in class deepnetts.data.ImageSet
-
Splits data set into several parts specified by the input parameter partSizes.
- split(double...) - Method in class deepnetts.data.TabularDataSet
-
Splits data set into several parts specified by the input parameter partSizes.
- split(int) - Method in class deepnetts.data.TabularDataSet
-
Split data set into specified number of part of equal sizes.
- sqr() - Method in class deepnetts.tensor.TensorBase
- sqrSumByCol(Tensor1D) - Method in class deepnetts.tensor.Tensor2D
- sqrt() - Method in class deepnetts.tensor.TensorBase
- standardDeviation(float[], float) - Static method in class deepnetts.util.Statistics
- Standardizer - Class in deepnetts.data.norm
-
Performs standardization on inputs in order to get desired statistical properties of the data set (zero mean and one standard deviation).
- Standardizer(DataSet) - Constructor for class deepnetts.data.norm.Standardizer
- STARTED - Enum constant in enum class deepnetts.net.train.TrainingEvent.Type
-
Specifies a training started event
- STARTED - Static variable in class deepnetts.net.train.TrainingEvent
-
Specifies that training has started.
- Statistics - Class in deepnetts.util
-
Statistical functions.
- Statistics() - Constructor for class deepnetts.util.Statistics
- Statistics.Histogram - Class in deepnetts.util
- Stats - Class in deepnetts.util
- Stats() - Constructor for class deepnetts.eval.ClassificationMetrics.Stats
- Stats() - Constructor for class deepnetts.util.Stats
- std(float) - Method in class deepnetts.tensor.TensorBase
- step(T) - Method in class deepnetts.automl.Range
- stop() - Method in class deepnetts.net.train.BackpropagationTrainer
-
Stops the training.
- STOP_EPOCHS - Static variable in class deepnetts.automl.Parameters
- STOP_ERROR - Static variable in class deepnetts.automl.Parameters
- STOPPED - Enum constant in enum class deepnetts.net.train.TrainingEvent.Type
-
Specifies a training stopped event
- STOPPED - Static variable in class deepnetts.net.train.TrainingEvent
-
Specifies that training has stopped.
- STRATCH - Enum constant in enum class deepnetts.util.ImageResize
- strategy(String) - Method in class deepnetts.automl.Range
- stride(int) - Method in class deepnetts.net.layers.Filter
- STRING - Enum constant in enum class deepnetts.util.ColumnType
- sub(float) - Method in class deepnetts.tensor.TensorBase
- sub(float[], float) - Static method in class deepnetts.tensor.Tensors
- sub(float[], float[]) - Static method in class deepnetts.tensor.Tensors
- sub(float, int, int) - Method in class deepnetts.tensor.Tensor2D
- sub(float, int, int, int) - Method in class deepnetts.tensor.Tensor3D
- sub(float, int, int, int, int) - Method in class deepnetts.tensor.Tensor4D
- sub(TensorBase) - Method in class deepnetts.tensor.TensorBase
-
Subtracts specified tensor t from this tensor.
- sub(TensorBase, TensorBase) - Method in class deepnetts.tensor.TensorBase
- subInplace(TensorBase, TensorBase) - Static method in class deepnetts.tensor.TensorBase
-
Subtracts tensor t2 from t1.
- submit(Callable) - Method in class deepnetts.util.DeepNettsThreadPool
-
Submit a single task to thread pool.
- sum() - Method in class deepnetts.tensor.TensorBase
- sumAbs() - Method in class deepnetts.tensor.TensorBase
-
Returns sum of abs values of this tensor - L1 norm
- sumByColInto(Tensor1D) - Method in class deepnetts.tensor.Tensor2D
- sumSqr() - Method in class deepnetts.tensor.TensorBase
-
Returns sum of squared values of this tensor - L2 norm
T
- TabularDataSet<T> - Class in deepnetts.data
-
Basic data set with tabular data.
- TabularDataSet() - Constructor for class deepnetts.data.TabularDataSet
- TabularDataSet(int, int) - Constructor for class deepnetts.data.TabularDataSet
-
Create a new instance of BasicDataSet with specified size of input and output.
- tanh(Tensor) - Method in interface deepnetts.accl.spi.ActivationVectorizationProvider
- Tanh - Class in deepnetts.net.layers.activation
-
Hyperbolic tangens activation function
- Tanh() - Constructor for class deepnetts.net.layers.activation.Tanh
- TANH - Enum constant in enum class deepnetts.net.layers.activation.ActivationType
- Tensor - Interface in deepnetts.tensor
-
A wrapper for multidimensional array/tensor.
- Tensor1D - Class in deepnetts.tensor
-
One dimensional tensor - a vector or an array.
- Tensor1D(float...) - Constructor for class deepnetts.tensor.Tensor1D
-
Creates a single row tensor with specified values.
- Tensor1D(int) - Constructor for class deepnetts.tensor.Tensor1D
- Tensor1D(Shape, float[]) - Constructor for class deepnetts.tensor.Tensor1D
- Tensor2D - Class in deepnetts.tensor
-
A 2D tensor / matrix with specified number of rows and columns..
- Tensor2D(float[][]) - Constructor for class deepnetts.tensor.Tensor2D
-
Creates a 2D tensor / matrix filled with given 2d array.
- Tensor2D(int, int) - Constructor for class deepnetts.tensor.Tensor2D
-
Creates a tensor with specified number of rows and columns (matrix).
- Tensor2D(int, int, float[]) - Constructor for class deepnetts.tensor.Tensor2D
- Tensor3D - Class in deepnetts.tensor
-
A 3D tensor/matrix, with rows, columns and depth.
- Tensor3D(float[][][]) - Constructor for class deepnetts.tensor.Tensor3D
-
Creates a 3D tensor from specified 3D array
- Tensor3D(int, int, int) - Constructor for class deepnetts.tensor.Tensor3D
-
Creates a new 3D tensor with specified number of rows, columns and depth.
- Tensor3D(int, int, int, float[]) - Constructor for class deepnetts.tensor.Tensor3D
- Tensor4D - Class in deepnetts.tensor
- Tensor4D(float[][][][]) - Constructor for class deepnetts.tensor.Tensor4D
- Tensor4D(int, int, int, int) - Constructor for class deepnetts.tensor.Tensor4D
- Tensor4D(int, int, int, int, float[]) - Constructor for class deepnetts.tensor.Tensor4D
- TensorBase - Class in deepnetts.tensor
-
The base call for multidimensional array/tensors.
- TensorBase(Shape) - Constructor for class deepnetts.tensor.TensorBase
- TensorBase(Shape, float...) - Constructor for class deepnetts.tensor.TensorBase
- TensorBase(TensorBase) - Constructor for class deepnetts.tensor.TensorBase
-
Public deep copy / clone constructor.
- TensorflowUtils - Class in deepnetts.util
- TensorflowUtils() - Constructor for class deepnetts.util.TensorflowUtils
- Tensors - Class in deepnetts.tensor
-
Static utility methods for tensors.
- TensorVectorizationProvider - Interface in deepnetts.accl.spi
- TensorVectorizationService - Class in deepnetts.accl.spi
- TensorVectorizationService() - Constructor for class deepnetts.accl.spi.TensorVectorizationService
- test(DataSet) - Method in class deepnetts.net.NeuralNetwork
-
Tests how good are predictions of this network using specified test set.
- TEST_FILE - Static variable in class deepnetts.util.ImageSetUtils
- testSet(DataSet) - Method in class deepnetts.automl.HyperParameterSearch
- threadPool - Variable in class deepnetts.net.layers.AbstractLayer
- threadPool - Variable in class deepnetts.net.NeuralNetwork
- toFile(String) - Method in class deepnetts.tensor.Tensor1D
- toFile(String) - Method in class deepnetts.tensor.Tensor2D
- toFile(String) - Method in class deepnetts.tensor.Tensor3D
- toFile(String) - Method in class deepnetts.tensor.Tensor4D
- toJson(NeuralNetwork) - Static method in class deepnetts.util.FileIO
-
Returns JSON representation of specified neural network object.
- Tokenizer - Interface in deepnetts.tokenizers
- toString() - Method in class deepnetts.automl.Parameter
- toString() - Method in class deepnetts.automl.Parameters.ParameterCombination
- toString() - Method in class deepnetts.automl.Parameters
- toString() - Method in class deepnetts.data.DataSetItem
- toString() - Method in class deepnetts.eval.ClassificationMetrics.Stats
- toString() - Method in class deepnetts.eval.ClassificationMetrics
- toString() - Method in class deepnetts.eval.ClassifierEvaluator
- toString() - Method in class deepnetts.eval.ConfusionMatrix
- toString() - Method in class deepnetts.eval.RegressionMetrics
- toString() - Method in class deepnetts.net.layers.CBOWEmbeddingLayer
- toString() - Method in class deepnetts.net.layers.ConvolutionalLayer
- toString() - Method in class deepnetts.net.layers.EmbeddingLayer
- toString() - Method in class deepnetts.net.layers.FlatEmbeddingLayer
- toString() - Method in class deepnetts.net.layers.FlattenLayer
- toString() - Method in class deepnetts.net.layers.FullyConnectedLayer
- toString() - Method in class deepnetts.net.layers.InputLayer
- toString() - Method in enum class deepnetts.net.layers.LayerType
- toString() - Method in class deepnetts.net.layers.MaxPoolingLayer
- toString() - Method in class deepnetts.net.layers.OutputLayer
- toString() - Method in class deepnetts.net.layers.SkipGramEmbeddingLayer
- toString() - Method in enum class deepnetts.net.loss.LossType
- toString() - Method in enum class deepnetts.net.NetworkType
- toString() - Method in class deepnetts.net.NeuralNetwork
-
Returns string representation of this network including all layers and settings.
- toString() - Method in class deepnetts.net.train.TrainingResult
- toString() - Method in class deepnetts.tensor.Shape
- toString() - Method in class deepnetts.tensor.Tensor3D
- toString() - Method in class deepnetts.tensor.TensorBase
- toString() - Method in record class deepnetts.tokenizers.bpe.Pair
-
Returns a string representation of this record class.
- toString() - Method in class deepnetts.util.BoundingBox
- toString() - Method in class deepnetts.util.CsvFormat
- toString() - Method in class deepnetts.util.LabelProbabilities
- toString() - Method in class deepnetts.util.Parameter
- toString() - Method in class deepnetts.util.Stats
- toTsr() - Method in class deepnetts.tensor.TensorBase
- train(String) - Static method in class deepnetts.tokenizers.WordTokenizer
- train(DataSet, double) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Run training using given training set, and split part of it to use as a validation set.
- train(DataSet) - Method in class deepnetts.net.NeuralNetwork
-
Trains the neural network using specified training set.
- train(DataSet) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Runs training using specified training set.
- train(DataSet) - Method in interface deepnetts.net.train.Trainer
-
Trains this model using specified training set.
- train(DataSet, DataSet) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Runs training using given training and validation sets.
- TRAIN - Enum constant in enum class deepnetts.net.Mode
- TRAIN_FILE - Static variable in class deepnetts.util.ImageSetUtils
- trainable - Variable in class deepnetts.net.layers.AbstractLayer
- Trainer - Interface in deepnetts.net.train
-
Generic interface for deep learning training algorithm.
- TrainerProvider<T> - Interface in deepnetts.net.train
-
This interface is implemented by trainable deep learning models, in order to provide access to training algorithm.
- TrainingEvent - Class in deepnetts.net.train
-
TrainingEvent is used to notify interested parties that training event has happened.
- TrainingEvent(BackpropagationTrainer, TrainingEvent.Type) - Constructor for class deepnetts.net.train.TrainingEvent
-
Constructs a new TrainingEvent with specified source and type.
- TrainingEvent.Type - Enum Class in deepnetts.net.train
-
Type of a training event.
- trainingListener(TrainingListener) - Method in class deepnetts.automl.HyperParameterSearch
- trainingListener(TrainingListener) - Method in class deepnetts.net.train.KFoldCrossValidation.Builder
- TrainingListener - Interface in deepnetts.net.train
-
The listener interface for receiving notifications about training events.
- TrainingResult - Class in deepnetts.net.train
-
All information about the completed training including training settings, epochs, loss and evaluation metrics.
- TrainingResult(Properties, int, float, EvaluationMetrics) - Constructor for class deepnetts.net.train.TrainingResult
- trainingSet(DataSet) - Method in class deepnetts.automl.HyperParameterSearch
- trainingSet(DataSet) - Method in class deepnetts.net.train.KFoldCrossValidation.Builder
- trainTestSplit(double) - Method in class deepnetts.data.TabularDataSet
- trainTestSplit(DataSet, double) - Static method in class deepnetts.data.DataSets
- TrainTestSplit - Class in deepnetts.data
-
This class holds training and test data set pair.
- TrainTestSplit(DataSet, DataSet) - Constructor for class deepnetts.data.TrainTestSplit
- translateImage(BufferedImage) - Static method in class deepnetts.util.ImageUtils
-
Generates 4 translated variations of the given image: translates it for one third of an image width and height in each direction.
- transposeInto(Tensor2D) - Method in class deepnetts.tensor.Tensor2D
- TRUE_NEGATIVE - Static variable in class deepnetts.eval.ConfusionMatrix
-
A label for items classified as negative which are really negative.
- TRUE_POSITIVE - Static variable in class deepnetts.eval.ConfusionMatrix
-
A label for items classified as positive which are really positive.
- TypedProperties - Class in deepnetts.util
-
Provides methods for getting typed properties for the given key.
- TypedProperties(Properties) - Constructor for class deepnetts.util.TypedProperties
U
- uniform(float[], float, float) - Static method in class deepnetts.net.weights.RandomWeights
- uniform(float[], int) - Static method in class deepnetts.net.weights.RandomWeights
-
Uniform U[-a,a] where a=1/sqrt(in).
- UNIFORM - Enum constant in enum class deepnetts.net.weights.RandomWeightsType
- updateLearningRate(float) - Method in class deepnetts.net.train.BackpropagationTrainer
-
Updates learning rate for all layers during the learning rate decay.
- useCuda() - Method in class deepnetts.core.DeepNetts
- useVectorAPI() - Method in class deepnetts.core.DeepNetts
V
- valueFor(NeuralNetwork, DataSet) - Method in interface deepnetts.net.loss.LossFunction
-
Calculates and returns loss function value for the given neural network and data set.
- valueOf(String) - Static method in enum class deepnetts.automl.SearchStrategy
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.layers.activation.ActivationType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.layers.LayerType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.loss.LossType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.Mode
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.NetworkType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.train.opt.OptimizerType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.train.TrainingEvent.Type
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.net.weights.RandomWeightsType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.util.ColumnType
-
Returns the enum constant of this class with the specified name.
- valueOf(String) - Static method in enum class deepnetts.util.ImageResize
-
Returns the enum constant of this class with the specified name.
- values - Variable in class deepnetts.tensor.TensorBase
-
Values stored in this tensor make it final , only input layer and tests sets values
- values() - Method in class deepnetts.automl.Parameters
- values() - Static method in enum class deepnetts.automl.SearchStrategy
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.layers.activation.ActivationType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.layers.LayerType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.loss.LossType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.Mode
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.NetworkType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.train.opt.OptimizerType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.train.TrainingEvent.Type
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.net.weights.RandomWeightsType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.util.ColumnType
-
Returns an array containing the constants of this enum class, in the order they are declared.
- values() - Static method in enum class deepnetts.util.ImageResize
-
Returns an array containing the constants of this enum class, in the order they are declared.
- valuesAsString(TensorBase[]) - Static method in class deepnetts.tensor.TensorBase
- variance(float[], float) - Static method in class deepnetts.util.Statistics
- vectorizationImpl - Static variable in class deepnetts.tensor.TensorBase
- version() - Method in class deepnetts.core.DeepNetts
-
Returns the Deep Netts version.
W
- weights - Variable in class deepnetts.net.layers.AbstractLayer
-
Input weight matrix / connectivity matrix for previous layer.
- WIDDROW_HOFF - Enum constant in enum class deepnetts.net.weights.RandomWeightsType
- widrowHoff(float[], float, float) - Static method in class deepnetts.net.weights.RandomWeights
- width - Variable in class deepnetts.net.layers.AbstractLayer
- wIdx - Variable in class deepnetts.tensor.Layout
- WordTokenizer - Class in deepnetts.tokenizers
- WordTokenizer(WordTokenizerParameters) - Constructor for class deepnetts.tokenizers.WordTokenizer
- WordTokenizerParameters - Class in deepnetts.tokenizers
- WordTokenizerParameters(Map, Map) - Constructor for class deepnetts.tokenizers.WordTokenizerParameters
- writeImages(List, String, String, String) - Static method in class deepnetts.util.ImageUtils
-
Writes list of images to specified file path.
- writeToFile(NeuralNetwork, String) - Static method in class deepnetts.util.FileIO
-
Serializes specified neural network to file with specified file.
- writeToFile(List, String) - Static method in class deepnetts.util.ImageSetUtils
-
Writes a given list of strings to file.
- writeToFileAsJson(NeuralNetwork, String) - Static method in class deepnetts.util.FileIO
X
- xavier(float[], int, int) - Static method in class deepnetts.net.weights.RandomWeights
-
Normalized uniform initialization U[-a,a] with a = sqrt(6/(in + out)).
- XAVIER - Enum constant in enum class deepnetts.net.weights.RandomWeightsType
Z
- zeroMean() - Method in class deepnetts.data.ImageSet
-
Applies zero mean normalization to entire dataset, and returns mean tensor.
- zeroMeanAndNormalize() - Method in class deepnetts.data.ImageSet
- zeroMeanPerChannel() - Method in class deepnetts.data.ImageSet
- zeros(int) - Static method in class deepnetts.tensor.Tensors
All Classes and Interfaces|All Packages|Constant Field Values|Serialized Form