mlrl.boosting.cython.learner module

@author: Michael Rapp (michael.rapp.ml@gmail.com)

class mlrl.boosting.cython.learner.AutomaticFeatureBinningMixin

Bases: ABC

Allows to configure a rule learner to automatically decide whether a method for the assignment of numerical feature values to bins should be used or not.

abstractmethod use_automatic_feature_binning()

Configures the rule learning to automatically decide whether a method for the assignment of numerical feature values to bins should be used or not.

class mlrl.boosting.cython.learner.AutomaticHeadMixin

Bases: ABC

Allows to configure a rule learner to automatically decide for the type of rule heads that should be used.

abstractmethod use_automatic_heads()

Configures the rule learner to automatically decide for the type of rule heads to be used.

class mlrl.boosting.cython.learner.AutomaticParallelRuleRefinementMixin

Bases: ABC

Allows to configure a rule learner to automatically decide whether multi-threading should be used for the parallel refinement of rules or not.

abstractmethod use_automatic_parallel_rule_refinement()

Configures the rule learner to automatically decide whether multi-threading should be used for the parallel refinement of rules or not.

class mlrl.boosting.cython.learner.AutomaticParallelStatisticUpdateMixin

Bases: ABC

Allows to configure a rule learner to automatically decide whether multi-threading should be used for the parallel update of statistics or not.

abstractmethod use_automatic_parallel_statistic_update()

Configures the rule learner to automatically decide whether multi-threading should be used for the parallel update of statistics or not.

class mlrl.boosting.cython.learner.CompleteHeadMixin

Bases: ABC

Allows to configure a rule learner to induce rules with complete heads that predict for all available outputs.

abstractmethod use_complete_heads()

Configures the rule learner to induce rules with complete heads that predict for all available outputs.

class mlrl.boosting.cython.learner.ConstantShrinkageMixin

Bases: ABC

Allows to configure a rule learner to use a post processor that shrinks the weights fo rules by a constant “shrinkage” parameter.

abstractmethod use_constant_shrinkage_post_processor() → ConstantShrinkageConfig

Configures the rule learner to use a post-processor that shrinks the weights of rules by a constant “shrinkage” parameter.

Returns:

A ConstantShrinkageConfig that allows further configuration of the post-processor

class mlrl.boosting.cython.learner.DecomposableSquaredErrorLossMixin

Bases: ABC

Allows to configure a rule learner to use a loss function that implements a multivariate variant of the squared error loss that is decomposable.

abstractmethod use_decomposable_squared_error_loss()

Configures the rule learner to use a loss function that implements a multivariate variant of the squared error loss that is decomposable.

class mlrl.boosting.cython.learner.DynamicPartialHeadMixin

Bases: ABC

Allows to configure a rule learner to induce rules with partial heads that predict for a subset of the available outputs that is determined dynamically.

abstractmethod use_dynamic_partial_heads() → DynamicPartialHeadConfig

Configures the rule learner to induce rules with partial heads that predict for a subset of the available outputs that is determined dynamically. Only those outputs for which the square of the predictive quality exceeds a certain threshold are included in a rule head.

Returns:

A DynamicPartialHeadConfig that allows further configuration of the rule heads

class mlrl.boosting.cython.learner.FixedPartialHeadMixin

Bases: ABC

Allows to configure a rule learner to induce rules with partial heads that predict for a predefined number of outputs.

abstractmethod use_fixed_partial_heads() → FixedPartialHeadConfig

Configures the rule learner to induce rules with partial heads that predict for a predefined number of outputs.

Returns:

A FixedPartialHeadConfig that allows further configuration of the rule heads

class mlrl.boosting.cython.learner.Float32StatisticsMixin

Bases: ABC

Allows to configure a rule learner to use 32-bit floating point values for representing gradients and Hessians.

abstractmethod use_32_bit_statistics()

Configures the rule learner to use 32-bit floating point values for representing gradients and Hessians.

class mlrl.boosting.cython.learner.Float64StatisticsMixin

Bases: ABC

Allows to configure a rule learner to use 64-bit floating point values for representing gradients and Hessians.

abstractmethod use_64_bit_statistics()

Configures the rule learner to use 64-bit floating point values for representing gradients and Hessians.

class mlrl.boosting.cython.learner.L1RegularizationMixin

Bases: ABC

Allows to configure a rule learner to use L1 regularization.

abstractmethod use_l1_regularization() → ManualRegularizationConfig

Configures the rule learner to use L1 regularization.

Returns:

A ManualRegularizationConfig that allows further configuration of the regularization term

class mlrl.boosting.cython.learner.L2RegularizationMixin

Bases: ABC

Allows to configure a rule learner to use L2 regularization.

abstractmethod use_l2_regularization() → ManualRegularizationConfig

Configures the rule learner to use L2 regularization.

Returns:

A ManualRegularizationConfig that allows further configuration of the regularization term

class mlrl.boosting.cython.learner.NoL1RegularizationMixin

Bases: ABC

Allows to configure a rule learner to not use L1 regularization.

abstractmethod use_no_l1_regularization()

Configures the rule learner to not use L1 regularization.

class mlrl.boosting.cython.learner.NoL2RegularizationMixin

Bases: ABC

Allows to configure a rule learner to not use L2 regularization.

abstractmethod use_no_l2_regularization()

Configures the rule learner to not use L2 regularization.

class mlrl.boosting.cython.learner.NonDecomposableSquaredErrorLossMixin

Bases: ABC

Allows to configure a rule learner to use a loss function that implements a multivariate variant of the squared error loss that is non-decomposable.

abstractmethod use_non_decomposable_squared_error_loss()

Configures the rule learner to use a loss function that implements a multivariant variant of the squared error loss that is non-decomposable.

class mlrl.boosting.cython.learner.OutputWiseScorePredictorMixin

Bases: ABC

Allows to configure a rule learner to use a predictor that predicts output-wise scores for given query examples by summing up the scores that are provided by individual rules for each output individually.

abstractmethod use_output_wise_score_predictor()

Configures the rule learner to use a predictor that predict output-wise scores for given query examples by summing up the scores that are provided by individual rules for each output individually.

class mlrl.boosting.cython.learner.SingleOutputHeadMixin

Bases: ABC

Allows to configure a rule learner to induce rules with single-output heads that predict for a single output.

abstractmethod use_single_output_heads()

Configures the rule learner to induce rules with single-output heads that predict for a single output.