Expand description
Multi-output regression and classification
This module provides meta-estimators for multi-target prediction problems. It includes strategies for independent multi-output prediction.
Re-exports§
pub use core::MultiOutputClassifier;pub use core::MultiOutputClassifierTrained;pub use core::MultiOutputRegressor;pub use core::MultiOutputRegressorTrained;pub use chains::BayesianClassifierChain;pub use chains::BayesianClassifierChainTrained;pub use chains::ChainMethod;pub use chains::ClassifierChain;pub use chains::ClassifierChainTrained;pub use chains::EnsembleOfChains;pub use chains::EnsembleOfChainsTrained;pub use chains::RegressorChain;pub use chains::RegressorChainTrained;pub use ensemble::GradientBoostingMultiOutput;pub use ensemble::GradientBoostingMultiOutputTrained;pub use ensemble::WeakLearner;pub use neural::ActivationFunction;pub use neural::AdversarialMultiTaskNetwork;pub use neural::AdversarialMultiTaskNetworkTrained;pub use neural::AdversarialStrategy;pub use neural::CellType;pub use neural::GradientReversalConfig;pub use neural::LambdaSchedule;pub use neural::LossFunction;pub use neural::MultiOutputMLP;pub use neural::MultiOutputMLPClassifier;pub use neural::MultiOutputMLPRegressor;pub use neural::MultiOutputMLPTrained;pub use neural::MultiTaskNeuralNetwork;pub use neural::MultiTaskNeuralNetworkTrained;pub use neural::RecurrentNeuralNetwork;pub use neural::RecurrentNeuralNetworkTrained;pub use neural::SequenceMode;pub use neural::TaskBalancing;pub use neural::TaskDiscriminator;pub use adversarial::AdversarialConfig;pub use regularization::GroupLasso;pub use regularization::GroupLassoTrained;pub use regularization::MetaLearningMultiTask;pub use regularization::MetaLearningMultiTaskTrained;pub use regularization::MultiTaskElasticNet;pub use regularization::MultiTaskElasticNetTrained;pub use regularization::NuclearNormRegression;pub use regularization::NuclearNormRegressionTrained;pub use regularization::RegularizationStrategy;pub use regularization::TaskClusteringRegressionTrained;pub use regularization::TaskClusteringRegularization;pub use regularization::TaskRelationshipLearning;pub use regularization::TaskRelationshipLearningTrained;pub use regularization::TaskSimilarityMethod;pub use correlation::CITestMethod;pub use correlation::CITestResult;pub use correlation::CITestResults;pub use correlation::ConditionalIndependenceTester;pub use correlation::CorrelationAnalysis;pub use correlation::CorrelationType;pub use correlation::DependencyGraph;pub use correlation::DependencyGraphBuilder;pub use correlation::DependencyMethod;pub use correlation::GraphStatistics;pub use correlation::OutputCorrelationAnalyzer;pub use transfer_learning::ContinualLearning;pub use transfer_learning::ContinualLearningTrained;pub use transfer_learning::CrossTaskTransferLearning;pub use transfer_learning::CrossTaskTransferLearningTrained;pub use transfer_learning::DomainAdaptation;pub use transfer_learning::DomainAdaptationTrained;pub use transfer_learning::KnowledgeDistillation;pub use transfer_learning::KnowledgeDistillationTrained;pub use transfer_learning::ProgressiveTransferLearning;pub use transfer_learning::ProgressiveTransferLearningTrained;pub use optimization::JointLossConfig;pub use optimization::JointLossOptimizer;pub use optimization::JointLossOptimizerTrained;pub use optimization::LossCombination;pub use optimization::LossFunction as OptimizationLossFunction;pub use optimization::MultiObjectiveConfig;pub use optimization::MultiObjectiveOptimizer;pub use optimization::MultiObjectiveOptimizerTrained;pub use optimization::NSGA2Algorithm;pub use optimization::NSGA2Config;pub use optimization::NSGA2Optimizer;pub use optimization::NSGA2OptimizerTrained;pub use optimization::ParetoSolution;pub use optimization::ScalarizationConfig;pub use optimization::ScalarizationMethod;pub use optimization::ScalarizationOptimizer;pub use optimization::ScalarizationOptimizerTrained;pub use probabilistic::BayesianMultiOutputConfig;pub use probabilistic::BayesianMultiOutputModel;pub use probabilistic::BayesianMultiOutputModelTrained;pub use probabilistic::EnsembleBayesianConfig;pub use probabilistic::EnsembleBayesianModel;pub use probabilistic::EnsembleBayesianModelTrained;pub use probabilistic::EnsembleStrategy;pub use probabilistic::GaussianProcessMultiOutput;pub use probabilistic::GaussianProcessMultiOutputTrained;pub use probabilistic::InferenceMethod;pub use probabilistic::KernelFunction;pub use probabilistic::PosteriorDistribution;pub use probabilistic::PredictionWithUncertainty;pub use probabilistic::PriorDistribution;pub use ranking::BinaryClassifierModel;pub use ranking::IndependentLabelPrediction;pub use ranking::IndependentLabelPredictionTrained;pub use ranking::ThresholdStrategy as RankingThresholdStrategy;pub use sparse_storage::sparse_utils;pub use sparse_storage::CSRMatrix;pub use sparse_storage::MemoryUsage;pub use sparse_storage::SparseMultiOutput;pub use sparse_storage::SparseMultiOutputTrained;pub use sparse_storage::SparsityAnalysis;pub use sparse_storage::StorageRecommendation;pub use streaming::IncrementalMultiOutputRegression;pub use streaming::IncrementalMultiOutputRegressionConfig;pub use streaming::IncrementalMultiOutputRegressionTrained;pub use streaming::StreamingMultiOutput;pub use streaming::StreamingMultiOutputConfig;pub use streaming::StreamingMultiOutputTrained;pub use performance::EarlyStopping;pub use performance::EarlyStoppingConfig;pub use performance::PredictionCache;pub use performance::WarmStartRegressor;pub use performance::WarmStartRegressorConfig;pub use performance::WarmStartRegressorTrained;pub use multi_label::BinaryRelevance;pub use multi_label::BinaryRelevanceTrained;pub use multi_label::LabelPowerset;pub use multi_label::LabelPowersetTrained;pub use multi_label::OneVsRestClassifier;pub use multi_label::OneVsRestClassifierTrained;pub use multi_label::PrunedLabelPowerset;pub use multi_label::PrunedLabelPowersetTrained;pub use multi_label::PruningStrategy;pub use tree::ClassificationCriterion;pub use tree::DAGInferenceMethod;pub use tree::MultiTargetDecisionTreeClassifier;pub use tree::MultiTargetDecisionTreeClassifierTrained;pub use tree::MultiTargetRegressionTree;pub use tree::MultiTargetRegressionTreeTrained;pub use tree::RandomForestMultiOutput;pub use tree::RandomForestMultiOutputTrained;pub use tree::TreeStructuredPredictor;pub use tree::TreeStructuredPredictorTrained;pub use neighbors::IBLRTrained;pub use neighbors::WeightFunction;pub use neighbors::IBLR;pub use svm::MLTSVMTrained;pub use svm::MultiOutputSVM;pub use svm::MultiOutputSVMTrained;pub use svm::RankSVM;pub use svm::RankSVMTrained;pub use svm::RankingSVMModel;pub use svm::SVMKernel;pub use svm::SVMModel;pub use svm::ThresholdStrategy as SVMThresholdStrategy;pub use svm::TwinSVMModel;pub use svm::MLTSVM;pub use sequence::FeatureFunction;pub use sequence::FeatureType;pub use sequence::HiddenMarkovModel;pub use sequence::HiddenMarkovModelTrained;pub use sequence::MaximumEntropyMarkovModel;pub use sequence::MaximumEntropyMarkovModelTrained;pub use sequence::StructuredPerceptron;pub use sequence::StructuredPerceptronTrained;pub use hierarchical::AggregationFunction;pub use hierarchical::ConsistencyEnforcement;pub use hierarchical::CostSensitiveHierarchicalClassifier;pub use hierarchical::CostSensitiveHierarchicalClassifierTrained;pub use hierarchical::CostStrategy;pub use hierarchical::GraphNeuralNetwork;pub use hierarchical::GraphNeuralNetworkTrained;pub use hierarchical::MessagePassingVariant;pub use hierarchical::OntologyAwareClassifier;pub use hierarchical::OntologyAwareClassifierTrained;pub use classification::CalibratedBinaryRelevance;pub use classification::CalibratedBinaryRelevanceTrained;pub use classification::CalibrationMethod;pub use classification::CostMatrix;pub use classification::CostSensitiveBinaryRelevance;pub use classification::CostSensitiveBinaryRelevanceTrained;pub use classification::DistanceMetric;pub use classification::MLkNN;pub use classification::MLkNNTrained;pub use classification::RandomLabelCombinations;pub use classification::SimpleBinaryModel;pub use metrics::average_precision_score;pub use metrics::confidence_interval;pub use metrics::coverage_error;pub use metrics::f1_score;pub use metrics::hamming_loss;pub use metrics::jaccard_score;pub use metrics::label_ranking_average_precision;pub use metrics::mcnemar_test;pub use metrics::one_error;pub use metrics::paired_t_test;pub use metrics::per_label_metrics;pub use metrics::precision_score_micro;pub use metrics::ranking_loss;pub use metrics::recall_score_micro;pub use metrics::subset_accuracy;pub use metrics::wilcoxon_signed_rank_test;pub use metrics::ConfidenceInterval;pub use metrics::PerLabelMetrics;pub use metrics::StatisticalTestResult;
Modules§
- activation
- Activation functions for neural networks
- adversarial
- Adversarial Multi-Task Networks with Feature Disentanglement
- chains
- Chain-based multi-output learning algorithms
- classification
- Multi-label classification algorithms
- core
- Core multi-output algorithms
- correlation
- Output Correlation Analysis and Dependency Modeling
- ensemble
- Ensemble methods for multi-output learning
- hierarchical
- Hierarchical classification and graph neural network models
- label_
analysis - Label combination frequency analysis utilities
- loss
- Loss functions for neural network training
- metrics
- Multi-output and multi-label evaluation metrics
- mlp
- Multi-Layer Perceptron for Multi-Output Learning
- multi_
label - Multi-label learning algorithms
- multitask
- Multi-Task Neural Networks with Shared Representation Learning
- neighbors
- Instance-based learning algorithms for multi-label classification
- neural
- Neural Network Multi-Output Learning
- optimization
- Multi-Output Learning Optimization Framework
- performance
- Performance Optimization for Multi-Output Learning
- probabilistic
- Auto-generated module structure
- ranking
- Label ranking and threshold optimization algorithms
- recurrent
- Neural Sequence Models for Structured Output Prediction
- regularization
- Multi-Task Regularization Methods
- sequence
- Sequence and structured prediction models
- sparse_
storage - Memory-efficient storage for sparse output representations
- streaming
- Streaming and Incremental Learning for Multi-Output Prediction
- svm
- Support Vector Machine algorithms for multi-output learning
- transfer_
learning - Transfer Learning for Multi-Task Learning
- tree
- Tree-based multi-output algorithms
- utilities
- Multi-label utilities and analysis tools
- utils
- Utility functions and shared types for multi-output learning