immuneML.hyperparameter_optimization package
Subpackages
- immuneML.hyperparameter_optimization.config package
- Submodules
- immuneML.hyperparameter_optimization.config.LeaveOneOutConfig module
- immuneML.hyperparameter_optimization.config.ManualSplitConfig module
- immuneML.hyperparameter_optimization.config.ReportConfig module
- immuneML.hyperparameter_optimization.config.SplitConfig module
- immuneML.hyperparameter_optimization.config.SplitType module
- Module contents
- immuneML.hyperparameter_optimization.core package
- immuneML.hyperparameter_optimization.states package
- Submodules
- immuneML.hyperparameter_optimization.states.HPAssessmentState module
- immuneML.hyperparameter_optimization.states.HPItem module
- immuneML.hyperparameter_optimization.states.HPLabelState module
- immuneML.hyperparameter_optimization.states.HPSelectionState module
- immuneML.hyperparameter_optimization.states.TrainMLModelState module
TrainMLModelState
TrainMLModelState.assessment
TrainMLModelState.assessment_states
TrainMLModelState.context
TrainMLModelState.dataset
TrainMLModelState.hp_settings
TrainMLModelState.hp_strategy
TrainMLModelState.label_configuration
TrainMLModelState.metrics
TrainMLModelState.name
TrainMLModelState.number_of_processes
TrainMLModelState.optimal_hp_item_paths
TrainMLModelState.optimal_hp_items
TrainMLModelState.optimization_metric
TrainMLModelState.path
TrainMLModelState.refit_optimal_model
TrainMLModelState.report_results
TrainMLModelState.reports
TrainMLModelState.selection
- Module contents
- immuneML.hyperparameter_optimization.strategy package
Submodules
immuneML.hyperparameter_optimization.HPSetting module
- class immuneML.hyperparameter_optimization.HPSetting.HPSetting(encoder: DatasetEncoder, encoder_params: dict, ml_method: MLMethod, ml_params: dict, preproc_sequence: list, encoder_name: str = None, ml_method_name: str = None, preproc_sequence_name: str = None)[source]
Bases:
object
immuneML.hyperparameter_optimization.HPSettingResult module
- class immuneML.hyperparameter_optimization.HPSettingResult.HPSettingResult(optimal_setting: HPSetting, all_settings: dict)[source]
Bases:
object
HPSettingResult encapsulates the results from evaluating a set of different hyperparameter settings (e.g. on one train/test split in the outer loop of nested cross-validation) - it stores the optimal setting which can be used to assess the performance on the task, and all settings if needed for downstream analysis.