API Reference

If you would like to use sklvq algorithms the most relevant part to look at is the “Predictors and Transformers” section. However, the other sections provide information about accepted parameters their range and default values.

Predictors and Transformers

LVQBaseClass([distance_type, ...])

Learning Vector Quantization base class

GLVQ([distance_type, distance_params, ...])

Generalized Learning Vector Quantization

GMLVQ([distance_type, distance_params, ...])

Generalized Matrix Learning Vector Quantization

LGMLVQ([distance_type, distance_params, ...])

Localized Generalized Matrix Learning Vector Quantization

Objective Functions

GeneralizedLearningObjective(...)

Generalized learning objective

Activation Functions

ActivationBaseClass()

Activation base class

Identity()

Identity function

Sigmoid([beta])

Sigmoid function

SoftPlus([beta])

Soft+ function

Swish([beta])

Swish function

Discriminant Functions

DiscriminantBaseClass()

Discriminant base class

RelativeDistance()

Relative distance function

Distance Functions

DistanceBaseClass()

Distance base class

Euclidean()

Euclidean distance function

SquaredEuclidean()

Squared Euclidean distance

AdaptiveSquaredEuclidean()

Adaptive squared Euclidean distance

LocalAdaptiveSquaredEuclidean()

Local adaptive squared Euclidean distance

Solvers

SolverBaseClass(objective)

Solver base class

SteepestGradientDescent(objective[, ...])

Steepest gradient descent (SGD)

WaypointGradientDescent(objective[, ...])

Waypoint gradient descent (WGD)

AdaptiveMomentEstimation(objective[, ...])

Adaptive moment estimation (ADAM)

BroydenFletcherGoldfarbShanno(objective, ...)

Broyden Fletcher Goldfarb Shanno (BFGS)

LimitedMemoryBfgs(objective, **kwargs)

Limited memory variant of BFGS (L-BFGS)