Package | Description |
---|---|
gov.sandia.cognition.learning.algorithm.gradient |
Provides gradient based learning algorithms.
|
gov.sandia.cognition.learning.algorithm.pca |
Provides implementations of Principle Components Analysis (PCA).
|
gov.sandia.cognition.learning.data.feature |
Provides data feature extractors.
|
gov.sandia.cognition.learning.function.kernel |
Provides kernel functions.
|
gov.sandia.cognition.learning.function.vector |
Provides functions that output vectors.
|
gov.sandia.cognition.math.matrix |
Provides interfaces and classes for linear algebra.
|
gov.sandia.cognition.math.signals |
Provides mathematical signal processing methods.
|
Modifier and Type | Interface and Description |
---|---|
interface |
GradientDescendable
Defines the functionality of an object that is required in order to apply
the gradient descent algorithm to it.
|
Modifier and Type | Class and Description |
---|---|
class |
GradientDescendableApproximator
Creates a
radientDescendable from a
VectorizableVectorFunction by estimating the parameter gradient
using a forward-difference approximation of the parameter Jacobian. |
Modifier and Type | Class and Description |
---|---|
class |
PrincipalComponentsAnalysisFunction
This VectorFunction maps a high-dimension input space onto a (hopefully)
simple low-dimensional output space by subtracting the mean of the input
data, and passing the zero-mean input through a dimension-reducing matrix
multiplication function.
|
Modifier and Type | Class and Description |
---|---|
class |
FeatureHashing
Implements a function that applies vector feature hashing.
|
Modifier and Type | Field and Description |
---|---|
protected VectorFunction |
VectorFunctionKernel.function
The vector function to use.
|
Modifier and Type | Method and Description |
---|---|
VectorFunction |
VectorFunctionKernel.getFunction()
Gets the vector function the kernel is using.
|
Modifier and Type | Method and Description |
---|---|
void |
VectorFunctionKernel.setFunction(VectorFunction function)
Sets the vector function for the kernel to use.
|
Constructor and Description |
---|
VectorFunctionKernel(VectorFunction function)
Creates a new VectorFunctionKernel from the given function.
|
VectorFunctionKernel(VectorFunction function,
Kernel<? super Vector> kernel)
Creates a new VectorFunctionKernel from the given function and kernel.
|
Modifier and Type | Class and Description |
---|---|
class |
DifferentiableFeedforwardNeuralNetwork
A feedforward neural network that can have an arbitrary number of layers,
and an arbitrary differentiable squashing (activation) function assigned to
each layer.
|
class |
DifferentiableGeneralizedLinearModel
A GradientDescenable version of a GeneralizedLinearModel, in
other words, a GeneralizedLinearModel where the squashing
function is differentiable
|
class |
ElementWiseDifferentiableVectorFunction
An ElementWiseVectorFunction that is also a DifferentiableVectorFunction
|
class |
ElementWiseVectorFunction
A VectorFunction that operates on each element of the Vector indepenently
of all others.
|
class |
FeedforwardNeuralNetwork
A feedforward neural network that can have an arbitrary number of layers,
and an arbitrary squashing (activation) function assigned to each layer.
|
class |
GaussianContextRecognizer
Uses a MixtureOfGaussians to compute the probability of the different
constituent MultivariateGaussians (that is, the contexts)
|
class |
GeneralizedLinearModel
A VectorizableVectorFunction that is a matrix multiply followed by a
VectorFunction...
|
class |
LinearCombinationVectorFunction
A weighted linear combination of scalar functions.
|
class |
LinearVectorFunction
The
LinearFunction class is a simple
VectorFunction that just scales the given input vector by a
scalar value. |
class |
MultivariateDiscriminant
Allows learning algorithms (vectorizing, differentiating) on a matrix*vector
multiply.
|
class |
MultivariateDiscriminantWithBias
A multivariate discriminant (matrix multiply) plus a constant vector
that gets added to the output of the discriminant.
|
class |
ThreeLayerFeedforwardNeuralNetwork
This is a "standard" feedforward neural network with a single hidden
layer.
|
Modifier and Type | Method and Description |
---|---|
VectorFunction |
GeneralizedLinearModel.getSquashingFunction()
Getter for squashingFunction
|
Modifier and Type | Method and Description |
---|---|
void |
GeneralizedLinearModel.setSquashingFunction(VectorFunction squashingFunction)
Setter for squashingFunction
|
Constructor and Description |
---|
GeneralizedLinearModel(MultivariateDiscriminant discriminant,
VectorFunction squashingFunction)
Creates a new instance of GeneralizedLinearModel
|
LinearCombinationVectorFunction(VectorFunction... basisFunctions)
Creates a new instance of LinearCombinationFunction
|
Modifier and Type | Interface and Description |
---|---|
interface |
DifferentiableVectorFunction
A VectorFunction that can is also differentiable
|
interface |
VectorizableDifferentiableVectorFunction
A VectorizableVectorFunction that also define a derivative
(this is needed for GradientDescendable).
|
interface |
VectorizableVectorFunction
The VectorizableVectorFunction interface defines a useful interface for
doing machine learning, which is a function that takes and returns vectors
and also is parameterizable as a vector.
|
Modifier and Type | Class and Description |
---|---|
class |
LinearDynamicalSystem
A generic Linear Dynamical System of the form
x_n = A*x_(n-1) + B*u_n y_n = C*x_n, where x_(n-1) is the previous state, x_n is the current state, u_n is the current input, y_n is the current output, A is the system matrix, B is the input-gain matrix, and C is the output-selector matrix |