A B C D E F G H I L M N O P R S T U V W

A

AbstractMinimizingOnlineLearner<M extends Model> - Class in de.jungblut.online.ml

AbstractMinimizingOnlineLearner(StochasticMinimizer) - Constructor for class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

AbstractOnlineLearner<M extends Model> - Class in de.jungblut.online.ml

AbstractOnlineLearner() - Constructor for class de.jungblut.online.ml.AbstractOnlineLearner

AdaptiveFTRLRegularizer - Class in de.jungblut.online.regularization

Based on the paper: http://www.eecs.tufts.edu/~dsculley/papers/ad-click-prediction.pdf

AdaptiveFTRLRegularizer(double, double, double) - Constructor for class de.jungblut.online.regularization.AdaptiveFTRLRegularizer

Creates a new AdaptiveFTRLRegularizer.

B

BayesianClassifier - Class in de.jungblut.online.bayes

BayesianClassifier(BayesianProbabilityModel) - Constructor for class de.jungblut.online.bayes.BayesianClassifier

BayesianProbabilityModel - Class in de.jungblut.online.bayes

BayesianProbabilityModel() - Constructor for class de.jungblut.online.bayes.BayesianProbabilityModel

BayesianProbabilityModel(DoubleMatrix, DoubleVector) - Constructor for class de.jungblut.online.bayes.BayesianProbabilityModel

breakOnDifference(double) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Breaks minimization process when the given delta in costs have been archieved.

build() - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

C

computeGradient(DoubleVector, DoubleVector, double, long, double, double) - Method in class de.jungblut.online.regularization.GradientDescentUpdater

computeGradient(DoubleVector, DoubleVector, double, long, double, double) - Method in class de.jungblut.online.regularization.L2Regularizer

computeGradient(DoubleVector, DoubleVector, double, long, double, double) - Method in interface de.jungblut.online.regularization.WeightUpdater

Computes the gradient.

computeMomentum() - Method in class de.jungblut.online.minimizer.StochasticGradientDescent

computeNewWeights(DoubleVector, DoubleVector, double, long, double, double) - Method in class de.jungblut.online.regularization.AdaptiveFTRLRegularizer

computeNewWeights(DoubleVector, DoubleVector, double, long, double, double) - Method in class de.jungblut.online.regularization.GradientDescentUpdater

Simplistic gradient descent without regularization.

computeNewWeights(DoubleVector, DoubleVector, double, long, double, double) - Method in class de.jungblut.online.regularization.L1Regularizer

computeNewWeights(DoubleVector, DoubleVector, double, long, double, double) - Method in interface de.jungblut.online.regularization.WeightUpdater

Computes the update for the given weights.

CostWeightTuple - Class in de.jungblut.online.regularization

CostWeightTuple(double, DoubleVector) - Constructor for class de.jungblut.online.regularization.CostWeightTuple

create(double) - Static method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Creates a new builder.

createModel(DoubleVector) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

Creates a model with the given minimized weights.

createModel(DoubleVector) - Method in class de.jungblut.online.regression.RegressionLearner

D

de.jungblut.classification.eval - package de.jungblut.classification.eval

de.jungblut.online.bayes - package de.jungblut.online.bayes

de.jungblut.online.minimizer - package de.jungblut.online.minimizer

de.jungblut.online.ml - package de.jungblut.online.ml

de.jungblut.online.regression - package de.jungblut.online.regression

de.jungblut.online.regression.multinomial - package de.jungblut.online.regression.multinomial

de.jungblut.online.regularization - package de.jungblut.online.regularization

deserialize(DataInput) - Method in class de.jungblut.online.bayes.BayesianProbabilityModel

deserialize(DataInput) - Method in interface de.jungblut.online.ml.Model

deserialize(DataInput) - Method in class de.jungblut.online.regression.multinomial.MultinomialRegressionModel

deserialize(DataInput) - Method in class de.jungblut.online.regression.RegressionModel

E

enableAdaptiveLearningRate() - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Enables adaptive learning rate, using the algorithm:
alpha = 1d / (initialAlpha * (allIterations + 2)); where allIterations is a counter over all passes.

ErrorCountingCallback - Class in de.jungblut.classification.eval

ErrorCountingCallback() - Constructor for class de.jungblut.classification.eval.ErrorCountingCallback

F

featureDimension - Variable in class de.jungblut.online.ml.AbstractOnlineLearner

G

getActivationFunction() - Method in class de.jungblut.online.regression.RegressionModel

getClassPriorProbability() - Method in class de.jungblut.online.bayes.BayesianProbabilityModel

getCost() - Method in class de.jungblut.online.regularization.CostWeightTuple

getModels() - Method in class de.jungblut.online.regression.multinomial.MultinomialRegressionModel

getProbabilityMatrix() - Method in class de.jungblut.online.bayes.BayesianProbabilityModel

getWeight() - Method in class de.jungblut.online.regularization.CostWeightTuple

getWeights() - Method in class de.jungblut.online.regression.RegressionModel

GradientDescentUpdater - Class in de.jungblut.online.regularization

GradientDescentUpdater() - Constructor for class de.jungblut.online.regularization.GradientDescentUpdater

H

historySize(int) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Sets the size of the history to keep to compute average improvements and output progress information.

holdoutValidationPercentage(double) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Holdout validation percentage, this will take a subset of the data on the stream and do a validation on it.

I

init(Supplier<Stream<FeatureOutcomePair>>) - Method in class de.jungblut.online.ml.AbstractOnlineLearner

IterationFinishedCallback - Interface in de.jungblut.online.minimizer

L

L1Regularizer - Class in de.jungblut.online.regularization

Ported to "real" Java from Spark's mllib org.apache.spark.mllib.optimization.Updater.

L1Regularizer() - Constructor for class de.jungblut.online.regularization.L1Regularizer

L2Regularizer - Class in de.jungblut.online.regularization

Computes the L2 regularized update: R(w) = (||w||^2) / 2.

L2Regularizer() - Constructor for class de.jungblut.online.regularization.L2Regularizer

lambda(double) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Sets the regularization parameter "lambda".

M

minimize(DoubleVector, Supplier<Stream<FeatureOutcomePair>>, StochasticCostFunction, int, boolean) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent

minimize(DoubleVector, Supplier<Stream<FeatureOutcomePair>>, StochasticCostFunction, int, boolean) - Method in interface de.jungblut.online.minimizer.StochasticMinimizer

Minimizes the given stochastic cost function on the supplied streams for the given amount of passes over the data.

minimizer - Variable in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

Model - Interface in de.jungblut.online.ml

momentum(double) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Add momentum to this gradient descent minimizer.

MultinomialRegressionClassifier - Class in de.jungblut.online.regression.multinomial

Classifier for multinomial regression.

MultinomialRegressionClassifier(MultinomialRegressionModel) - Constructor for class de.jungblut.online.regression.multinomial.MultinomialRegressionClassifier

Constructs a new multinomial regression classifier that does normalization over independent predictions.

MultinomialRegressionClassifier(MultinomialRegressionModel, boolean) - Constructor for class de.jungblut.online.regression.multinomial.MultinomialRegressionClassifier

Constructs a new multinomial regression classifier that does normalization over independent predictions by summing over the predictions and dividing each entry.

MultinomialRegressionLearner - Class in de.jungblut.online.regression.multinomial

A regression learner that learns multiple independent regression models and blends them into a single model.

MultinomialRegressionLearner(StochasticMinimizer, ActivationFunction, LossFunction) - Constructor for class de.jungblut.online.regression.multinomial.MultinomialRegressionLearner

MultinomialRegressionLearner(IntFunction<RegressionLearner>) - Constructor for class de.jungblut.online.regression.multinomial.MultinomialRegressionLearner

MultinomialRegressionModel - Class in de.jungblut.online.regression.multinomial

MultinomialRegressionModel() - Constructor for class de.jungblut.online.regression.multinomial.MultinomialRegressionModel

MultinomialRegressionModel(RegressionModel[]) - Constructor for class de.jungblut.online.regression.multinomial.MultinomialRegressionModel

N

NaiveBayesLearner - Class in de.jungblut.online.bayes

Multinomial naive bayes learner.

NaiveBayesLearner() - Constructor for class de.jungblut.online.bayes.NaiveBayesLearner

Default constructor to construct this classifier.

NaiveBayesLearner(boolean) - Constructor for class de.jungblut.online.bayes.NaiveBayesLearner

Pass true if this classifier should output some progress information to the logger.

numOutcomeClasses - Variable in class de.jungblut.online.ml.AbstractOnlineLearner

numPasses - Variable in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

O

observeExample(FeatureOutcomePair, DoubleVector) - Method in interface de.jungblut.online.minimizer.StochasticCostFunction

Observes the next example using the given weights.

observeExample(FeatureOutcomePair, DoubleVector) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

Observes the next example.

observeExample(FeatureOutcomePair, DoubleVector) - Method in class de.jungblut.online.regression.RegressionLearner

observeExampleSafe(FeatureOutcomePair, DoubleVector) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

onIterationFinished(int, long, double, DoubleVector, boolean) - Method in class de.jungblut.classification.eval.ErrorCountingCallback

onIterationFinished(int, long, double, DoubleVector, boolean) - Method in interface de.jungblut.online.minimizer.IterationFinishedCallback

This callback when a pass over an example in a stream of a minimization objective is finished.

OnlineLearner<M extends Model> - Interface in de.jungblut.online.ml

OnlineLearning interface.

onPassFinished(int, long, double, DoubleVector) - Method in class de.jungblut.classification.eval.ErrorCountingCallback

onPassFinished(int, long, double, DoubleVector) - Method in class de.jungblut.classification.eval.RegressionValidationCallback

onPassFinished(int, long, double, DoubleVector) - Method in interface de.jungblut.online.minimizer.PassFinishedCallback

This callback when a pass over a stream of a minimization objective is finished.

onValidationFinished(int, long, double, DoubleVector, FeatureOutcomePair) - Method in class de.jungblut.classification.eval.RegressionValidationCallback

onValidationFinished(int, long, double, DoubleVector, FeatureOutcomePair) - Method in interface de.jungblut.online.minimizer.ValidationFinishedCallback

This callback when a pass over an example in a stream of a minimization objective is finished.

outcomeDimension - Variable in class de.jungblut.online.ml.AbstractOnlineLearner

P

PassFinishedCallback - Interface in de.jungblut.online.minimizer

peekDimensions(Supplier<Stream<FeatureOutcomePair>>) - Method in class de.jungblut.online.ml.AbstractOnlineLearner

Peeks for the feature and outcome dimensions.

predict(DoubleVector) - Method in class de.jungblut.online.bayes.BayesianClassifier

predict(DoubleVector) - Method in class de.jungblut.online.regression.multinomial.MultinomialRegressionClassifier

predict(DoubleVector) - Method in class de.jungblut.online.regression.RegressionClassifier

progressReportInterval(int) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Sets the progress report interval.

R

random - Variable in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

randomInitialize(int) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

RegressionClassifier - Class in de.jungblut.online.regression

Classifier for regression model.

RegressionClassifier(RegressionModel) - Constructor for class de.jungblut.online.regression.RegressionClassifier

RegressionClassifier(DoubleVector, ActivationFunction) - Constructor for class de.jungblut.online.regression.RegressionClassifier

RegressionLearner - Class in de.jungblut.online.regression

A regression learner that learns weights on a stream, given an optimization objective (e.g. log loss).

RegressionLearner(StochasticMinimizer, ActivationFunction, LossFunction) - Constructor for class de.jungblut.online.regression.RegressionLearner

RegressionModel - Class in de.jungblut.online.regression

RegressionModel() - Constructor for class de.jungblut.online.regression.RegressionModel

RegressionModel(DoubleVector, ActivationFunction) - Constructor for class de.jungblut.online.regression.RegressionModel

RegressionValidationCallback - Class in de.jungblut.classification.eval

RegressionValidationCallback(RegressionLearner) - Constructor for class de.jungblut.classification.eval.RegressionValidationCallback

S

serialize(DataOutput) - Method in class de.jungblut.online.bayes.BayesianProbabilityModel

serialize(DataOutput) - Method in interface de.jungblut.online.ml.Model

serialize(DataOutput) - Method in class de.jungblut.online.regression.multinomial.MultinomialRegressionModel

serialize(DataOutput) - Method in class de.jungblut.online.regression.RegressionModel

setIterationCallback(IterationFinishedCallback) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent

setNumPasses(int) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

setPassCallback(PassFinishedCallback) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent

setRandom(Random) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

setValidationCallback(ValidationFinishedCallback) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent

setVerbose(boolean) - Method in class de.jungblut.online.ml.AbstractOnlineLearner

sparseWeights - Variable in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

StochasticCostFunction - Interface in de.jungblut.online.minimizer

StochasticGradientDescent - Class in de.jungblut.online.minimizer

Stochastic gradient descent.

StochasticGradientDescent.StochasticGradientDescentBuilder - Class in de.jungblut.online.minimizer

StochasticMinimizer - Interface in de.jungblut.online.minimizer

streamSupplier - Variable in class de.jungblut.online.ml.AbstractOnlineLearner

T

train(Supplier<Stream<FeatureOutcomePair>>) - Method in class de.jungblut.online.bayes.NaiveBayesLearner

train(Supplier<Stream<FeatureOutcomePair>>) - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

train(Supplier<Stream<FeatureOutcomePair>>) - Method in interface de.jungblut.online.ml.OnlineLearner

Trains a new model using the supplied streams.

train(Supplier<Stream<FeatureOutcomePair>>) - Method in class de.jungblut.online.regression.multinomial.MultinomialRegressionLearner

U

updateWeights(CostGradientTuple) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent

useSparseWeights() - Method in class de.jungblut.online.ml.AbstractMinimizingOnlineLearner

V

ValidationFinishedCallback - Interface in de.jungblut.online.minimizer

verbose - Variable in class de.jungblut.online.ml.AbstractOnlineLearner

verbose() - Method in class de.jungblut.online.ml.AbstractOnlineLearner

W

weightUpdater(WeightUpdater) - Method in class de.jungblut.online.minimizer.StochasticGradientDescent.StochasticGradientDescentBuilder

Sets the weight updater, for example to use regularization.

WeightUpdater - Interface in de.jungblut.online.regularization

A B C D E F G H I L M N O P R S T U V W