File learner_classification.hpp¶

class IClassificationRuleLearner¶

#include <learner_classification.hpp>

Defines an interface for all rule learners that can be applied to classification problems.

Subclassed by AbstractClassificationRuleLearner

Public Functions

inline virtual ~IClassificationRuleLearner()¶

virtual std::unique_ptr<ITrainingResult> fit(const IExampleWeights &exampleWeights, const IFeatureInfo &featureInfo, const IColumnWiseFeatureMatrix &featureMatrix, const IRowWiseLabelMatrix &labelMatrix) const = 0¶

Applies the rule learner to given training examples and corresponding ground truth labels.

Parameters:

exampleWeights – A reference to an object of type IExampleWeights that provides access to the weights of the training examples
featureInfo – A reference to an object of type IFeatureInfo that provides information about the types of individual features
featureMatrix – A reference to an object of type IColumnWiseFeatureMatrix that provides column-wise access to the feature values of the training examples
labelMatrix – A reference to an object of type IRowWiseLabelMatrix that provides row-wise access to the ground truth labels of the training examples

Returns:

An unique pointer to an object of type ITrainingResult that provides access to the results of fitting the rule learner to the training data

inline virtual bool canPredictScores(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Returns whether the rule learner is able to predict scores or not.

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

True, if the rule learner is able to predict scores, false otherwise

virtual bool canPredictScores(const IRowWiseFeatureMatrix &featureMatrix, uint32 numLabels) const = 0¶

Returns whether the rule learner is able to predict scores or not.

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
numLabels – The number of labels to predict for

Returns:

True, if the rule learner is able to predict scores, false otherwise

inline virtual std::unique_ptr<IScorePredictor> createScorePredictor(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Creates and returns a predictor that may be used to predict scores for given query examples. If the prediction of scores is not supported by the rule learner, a std::runtime_error is thrown.

Throws:

std::runtime_exception – The exception that is thrown if the prediction of scores is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

An unique pointer to an object of type IScorePredictor that may be used to predict scores for the given query examples

virtual std::unique_ptr<IScorePredictor> createScorePredictor(const IRowWiseFeatureMatrix &featureMatrix, const IRuleModel &ruleModel, const IOutputSpaceInfo &outputSpaceInfo, uint32 numLabels) const = 0¶

Creates and returns a predictor that may be used to predict scores for given query examples. If the prediction of scores is not supported by the rule learner, a std::runtime_error is thrown.

Throws:

std::runtime_exception – The exception that is thrown if the prediction of scores is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
ruleModel – A reference to an object of type IRuleModel that should be used to obtain predictions
outputSpaceInfo – A reference to an object of type IOutputSpaceInfo that provides information about the output space that may be used as a basis for obtaining predictions
numLabels – The number of labels to predict for

Returns:

An unique pointer to an object of type IScorePredictor that may be used to predict scores for the given query examples

inline virtual bool canPredictProbabilities(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Returns whether the rule learner is able to predict probabilities or not.

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

True, if the rule learner is able to predict probabilities, false otherwise

virtual bool canPredictProbabilities(const IRowWiseFeatureMatrix &featureMatrix, uint32 numLabels) const = 0¶

Returns whether the rule learner is able to predict probabilities or not.

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
numLabels – The number of labels to predict for

Returns:

True, if the rule learner is able to predict probabilities, false otherwise

inline virtual std::unique_ptr<IProbabilityPredictor> createProbabilityPredictor(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Creates and returns a predictor that may be used to predict probability estimates for given query examples. If the prediction of probability estimates is not supported by the rule learner, a std::runtime_error is thrown.

Throws:

std::runtime_exception – The exception that is thrown if the prediction of probability estimates is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

An unique pointer to an object of type IProbabilityPredictor that may be used to predict probability estimates for the given query examples

virtual std::unique_ptr<IProbabilityPredictor> createProbabilityPredictor(const IRowWiseFeatureMatrix &featureMatrix, const IRuleModel &ruleModel, const IOutputSpaceInfo &outputSpaceInfo, const IMarginalProbabilityCalibrationModel &marginalProbabilityCalibrationModel, const IJointProbabilityCalibrationModel &jointProbabilityCalibrationModel, uint32 numLabels) const = 0¶

Throws:

std::runtime_exception – The exception that is thrown if the prediction of probability estimates is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
ruleModel – A reference to an object of type IRuleModel that should be used to obtain predictions
outputSpaceInfo – A reference to an object of type IOutputSpaceInfo that provides information about the output space that may be used as a basis for obtaining predictions
marginalProbabilityCalibrationModel – A reference to an object of type IMarginalProbabilityCalibrationModel that may be used for the calibration of marginal probabilities
jointProbabilityCalibrationModel – A reference to an object of type IJointProbabilityCalibrationModel that may be used for the calibration of joint probabilities
numLabels – The number of labels to predict for

Returns:

An unique pointer to an object of type IProbabilityPredictor that may be used to predict probability estimates for the given query examples

inline virtual bool canPredictBinary(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Returns whether the rule learner is able to predict binary labels or not.

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

True, if the rule learner is able to predict binary labels, false otherwise

virtual bool canPredictBinary(const IRowWiseFeatureMatrix &featureMatrix, uint32 numLabels) const = 0¶

Returns whether the rule learner is able to predict binary labels or not.

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
numLabels – The number of labels to predict for

Returns:

True, if the rule learner is able to predict binary labels, false otherwise

inline virtual std::unique_ptr<IBinaryPredictor> createBinaryPredictor(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Creates and returns a predictor that may be used to predict binary labels for given query examples. If the prediction of binary labels is not supported by the rule learner, a std::runtime_error is thrown.

Throws:

std::runtime_exception – The exception that is thrown if the prediction of binary labels is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

An unique pointer to an object of type IBinaryPredictor that may be used to predict binary labels for the given query examples

virtual std::unique_ptr<IBinaryPredictor> createBinaryPredictor(const IRowWiseFeatureMatrix &featureMatrix, const IRuleModel &ruleModel, const IOutputSpaceInfo &outputSpaceInfo, const IMarginalProbabilityCalibrationModel &marginalProbabilityCalibrationModel, const IJointProbabilityCalibrationModel &jointProbabilityCalibrationModel, uint32 numLabels) const = 0¶

Throws:

std::runtime_exception – The exception that is thrown if the prediction of binary labels is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
ruleModel – A reference to an object of type IRuleModel that should be used to obtain predictions
outputSpaceInfo – A reference to an object of type IOutputSpaceInfo that provides information about the output space that may be used as a basis for obtaining predictions
marginalProbabilityCalibrationModel – A reference to an object of type IMarginalProbabilityCalibrationModel that may be used for the calibration of marginal probabilities
jointProbabilityCalibrationModel – A reference to an object of type IJointProbabilityCalibrationModel that may be used for the calibration of joint probabilities
numLabels – The number of labels to predict for

Returns:

An unique pointer to an object of type IBinaryPredictor that may be used to predict binary labels for the given query examples

inline virtual std::unique_ptr<ISparseBinaryPredictor> createSparseBinaryPredictor(const IRowWiseFeatureMatrix &featureMatrix, const ITrainingResult &trainingResult) const¶

Creates and returns a predictor that may be used to predict sparse binary labels for given query examples. If the prediction of sparse binary labels is not supported by the rule learner, a std::runtime_error is thrown.

Throws:

std::runtime_exception – The exception that is thrown if the prediction of sparse binary labels is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
trainingResult – A reference to an object of type ITrainingResult that provides access to the model and additional information that should be used to obtain predictions

Returns:

An unique pointer to an object of type ISparseBinaryPredictor that may be used to predict sparse binary labels for the given query examples

virtual std::unique_ptr<ISparseBinaryPredictor> createSparseBinaryPredictor(const IRowWiseFeatureMatrix &featureMatrix, const IRuleModel &ruleModel, const IOutputSpaceInfo &outputSpaceInfo, const IMarginalProbabilityCalibrationModel &marginalProbabilityCalibrationModel, const IJointProbabilityCalibrationModel &jointProbabilityCalibrationModel, uint32 numLabels) const = 0¶

Throws:

std::runtime_exception – The exception that is thrown if the prediction of sparse binary labels is not supported by the rule learner

Parameters:

featureMatrix – A reference to an object of type IRowWiseFeatureMatrix that provides row-wise access to the feature values of the query examples
ruleModel – A reference to an object of type IRuleModel that should be used to obtain predictions
outputSpaceInfo – A reference to an object of type IOutputSpaceInfo that provides information about the output space that may be used as a basis for obtaining predictions
marginalProbabilityCalibrationModel – A reference to an object of type IMarginalProbabilityCalibrationModel that may be used for the calibration of marginal probabilities
jointProbabilityCalibrationModel – A reference to an object of type IJointProbabilityCalibrationModel that may be used for the calibration of joint probabilities
numLabels – The number of labels to predict for

Returns:

An unique pointer to an object of type ISparseBinaryPredictor that may be used to predict sparse binary labels for the given query examples

class IOutputWiseStratifiedInstanceSamplingMixin : public virtual IRuleLearnerConfig ¶

#include <learner_classification.hpp>

Defines an interface for all classes that allow to configure a rule learner to use label-wise stratified instance sampling.

Public Functions

inline virtual ~IOutputWiseStratifiedInstanceSamplingMixin() override¶

inline virtual IOutputWiseStratifiedInstanceSamplingConfig &useOutputWiseStratifiedInstanceSampling()¶

Configures the rule learner to sample from the available training examples using stratification, such that for each label the proportion of relevant and irrelevant examples is maintained, whenever a new rule should be learned.

Returns:: A reference to an object of type IOutputWiseStratifiedInstanceSamplingConfig that allows further configuration of the method for sampling instances

class IExampleWiseStratifiedInstanceSamplingMixin : public virtual IRuleLearnerConfig ¶

#include <learner_classification.hpp>

Defines an interface for all classes that allow to configure a rule learner to use example-wise stratified instance sampling.

Public Functions

inline virtual ~IExampleWiseStratifiedInstanceSamplingMixin() override¶

inline virtual IExampleWiseStratifiedInstanceSamplingConfig &useExampleWiseStratifiedInstanceSampling()¶

Configures the rule learner to sample from the available training examples using stratification, where distinct label vectors are treated as individual classes, whenever a new rule should be learned.

Returns:: A reference to an object of type IExampleWiseStratifiedInstanceSamplingConfig that allows further configuration of the method for sampling instances

class IOutputWiseStratifiedBiPartitionSamplingMixin : public virtual IRuleLearnerConfig ¶

#include <learner_classification.hpp>

Defines an interface for all classes that allow to configure a rule learner to partition the available training examples into a training set and a holdout set using stratification, such that for each label the proportion of relevant and irrelevant examples is maintained.

Public Functions

inline virtual ~IOutputWiseStratifiedBiPartitionSamplingMixin() override¶

inline virtual IOutputWiseStratifiedBiPartitionSamplingConfig &useOutputWiseStratifiedBiPartitionSampling()¶

Configures the rule learner to partition the available training examples into a training set and a holdout set using stratification, such that for each label the proportion of relevant and irrelevant examples is maintained.

Returns:: A reference to an object of type IOutputWiseStratifiedBiPartitionSamplingConfig that allows further configuration of the method for partitioning the available training examples into a training and a holdout set

class IExampleWiseStratifiedBiPartitionSamplingMixin : public virtual IRuleLearnerConfig ¶

#include <learner_classification.hpp>

Defines an interface for all classes that allow to configure a rule learner to partition the available training examples into a training set and a holdout set using stratification, where distinct label vectors are treated as individual classes.

Public Functions

inline virtual ~IExampleWiseStratifiedBiPartitionSamplingMixin() override¶

inline virtual IExampleWiseStratifiedBiPartitionSamplingConfig &useExampleWiseStratifiedBiPartitionSampling()¶

Configures the rule learner to partition the available training examples into a training set and a holdout set using stratification, where distinct label vectors are treated as individual classes.

Returns:: A reference to an object of type IExampleWiseStratifiedBiPartitionSamplingConfig that allows further configuration of the method for partitioning the available training examples into a training and a holdout set