All Classes and Interfaces
Class
Description
Wraps a base
Tensor by traversing only its elements in a specified range (from begin, up to end-1).Implements an accuracy
Loss of row-by-row comparisons.Thic class implements an Adam
Optimizer as explained in the paper:
Kingma, Diederik P., and Jimmy Ba.Implements a
NNOperation that adds its two inputs.Extends the
ModelTraining class to be able to train
Model instances for attributed graph functions (AGFs).Demonstrates classification with an APPNP GNN.
Implements a
NNOperation that creates a version of adjacency matrices
with column-wise attention involving neighbor similarity.Wraps an
Optimizer by accumulating derivatives and calling
Optimizer.update(Tensor, Tensor) with the average derivative
after a fixed number of accumulations.Implements a binary cross-entropy
For more than one output dimensions use
Loss.For more than one output dimensions use
CategoricalCrossEntropyImplements a categorical cross-entropy
For binary classification of one output use
Loss.For binary classification of one output use
BinaryCrossEntropy.Downloads and constructs the Citeseer node classification
Dataset.Defines a matrix whose columns are all a copy of a
Tensor.Implements a
NNOperation that performs the operation 1-x for its
simple input x.Implements a
NNOperation that concatenates its two matrix inputs.Implements a
NNOperation that holds a constant tensor.Downloads and constructs the Cora node classification
Dataset.This class provides the backbone with which to define datasets.
Implements a dense
Matrix where all elements are stored in memory.This class provides a dense
Tensor that wraps an array of doubles.Implements a square matrix whose diagonal elements are determined by the
correspond values of an underlying tensor and off-diagonal elements are zero.
This interface abstracts a probability distribution that can be passed to
Tensor.setToRandom(Distribution) for random tensor initialization.Implements a
NNOperation that converts its first argument to a
ColumnRepetition matrix with a number of columns equal to the second
argument.A
Matrix without data that contains only the correct dimension names
and sizes.A
Tensor without data that contains only the correct dimension names
and sizes.Implements a
NNOperation that performs an element-by-element
exponential transformation of its one input tensor.Extends the capabilities of
LayeredBuilder to use for node
classification.Implements a
NNOperation that lists the first element of the 2D
matrix element iterator.Implements a
NNOperation that performs the equivalent of TensorFlow's
gather operation.Demonstrates classification with the GCN architecture.
Implements a gradient descent
Optimizer.Converts back-and-forth between objects and unique ids.
Implements a
NNOperation that just transfers its single input.This class defines an abstract interface for applying initializers to models.
This is a
VariancePreservingInitializer.This is a
VariancePreservingInitializer.Implements a
NNOperation that performs a L1 transformation of its one
input tensor by row or by column.Extends the capabilities of the
ModelBuilder with the ability to
define multilayer (e.g.This implementation covers code of the Learning tutorial.
Implements a
NNOperation that outputs the natural logarithm of its
single input.Demonstrates classification with logistic regression.
Provides computation and (partial) derivation of popular activation functions
and cross-entropy loss functions.
This class provides an abstract implementation of loss functions to be used
during
Model training.Implements a
NNOperation that performs a leaky relu operation, where
the first argument is a tensor on which it is applied and the second one
should be a tensor wrapping a double value (consider initializing this with
as a Constant holding a tensor generated with
Tensor.fromDouble(double)) where the wrapped value indicates the
negative region's slope.Implements a
NNOperation that multiplies its two matrix inputs.This class provides an abstract implementation of Matrix functionalities.
Implements a
NNOperation that performs row-wise or column-wise
maximum reduction on vector tensors or matrices.Implements a
NNOperation that performs row-wise or column-wise
mean reduction on vector tensors or matrices.A memory management system for thread-safe allocation and release of arrays
of doubles.
Demonstrates classification with a message passing architecture.
Demonstrates classification with a two-layer perceptron
This class is a way to organize
NNOperation trees into trainable
machine learning models.This class and subclasses can be used to create
Model instances by
automatically creating and managing NNOperation instances based on
textual descriptions.Demonstrates model builder internal node access that allows training with
a symbolically defined loss function.
This is a helper class that automates the definition of training processes of
Model instances by defining the number of epochs, loss functions,
number of batches and the ability to use ThreadPool for parallelized
batch computations.Implements a
NNOperation that multiplies its two inputs
element-by-element.Extends the base
ModelBuilder with the full capabilities of the
Neuralang scripting language.Implements a
NNOperation that performs an exponential transformation
of its single input, but only on the non-zero elements.This implementation covers code of the Neural Networks tutorial.
This class defines an abstract neural network operation with forward and
backpropagation capabilities.
Implements a Normal
Distribution of given mean and standard
deviation.Provides an interface for training tensors.
Implements a
NNOperation that holds and returns a parameter tensor.Downloads and constructs the Pubmed node classification
Dataset.Demonstrates classification with the GCN architecture.
Implements an iterator that traverses a range [min, max) where the right side
is non-inclusive.
Implements an iterator that traverses a two-dimensional range (min, max) x (min2, max2).
Wraps an
Optimizer by applying the derivative of L2 loss on every
tensor during Optimizer.update(Tensor, Tensor).Implements a
NNOperation that performs a relu transformation of its
one input tensor.Implements a
NNOperation that converts its first argument to a
ColumnRepetition matrix with a number of columns equal to the second
argument.Implements a
Matrix whose elements are all equals.This class provides
Tensor whose elements are all equal.Implements a
NNOperation that reshapes a matrix.Defines a matrix whose rows are all a copy of a
Tensor.Extends the
ModelTraining class to train Model
instances from feature and label matrices.Demonstrates classification with an architecture defined through the scripting engine.
Implements a
NNOperation that performs a sigmoid transformation of
its single input.Demonstrates custom initialization of parameters.
This class provices an interface with which to define data slices, for
instance to sample labels.
Implements a
NNOperation that performs row-wise or column-wise
softmax on vector tensors or matrices.A sparse
Matrix that allocates memory only for non-zero elements.Deprecated.
Under development.
This class provides a sparse
Tensor with many zero elements.Implements a
NNOperation that performs row-wise or column-wise
sum reduction on vector tensors or matrices.Implements a
NNOperation that performs a tanh transformation of its
single input.This class provides a native java implementation of Tensor functionalities.
This class provides thread execution pool utilities while keeping track of
thread identifiers for use by thread-specific
NNOperation.Implements a
NNOperation that lists the second element of the 2D
matrix element iterator.This class generates trajectory graph labels.
Implements a
NNOperation that performs matrix transposition.Generates a transposed version of a base matrix, with which it shares
elements.
Implements a Uniform
Distribution of given bounds.Implements a
NNOperation that represents Model inputs.This class describes a broad class of
Initializer strategies, in which
dense neural layer initialization is controlled so that variance is mostly preserved from
inputs to outputs to avoid vanishing or exploding gradients in the first training
runs.Implements a dense
Matrix where all elements are stored in memory.This class provides a dense
Tensor that wraps an array of doubles.Implements a
Loss that wraps other losses and outputs their value
during training to an output stream (to System.out by default).Wraps a list of tensors into a matrix with the tensors as columns.
Wraps a list of tensors into a matrix with the tensors as rows.
This is a
VariancePreservingInitializer.This is a
VariancePreservingInitializer.