Optimization#
- class HyperparameterTuning(method_name: str, description: str, plot_settings)[source]#
Bases:
MeasureEvaluatorPerform hyperparameter tuning using grid or random search.
This evaluator provides comprehensive hyperparameter optimization capabilities using either grid search or random search methods. It supports visualization of tuning results for 1D, 2D, and multi-dimensionalparameter spaces using line plots, 3D surface plots, and parallel coordinate plots respectively.
- Attributes:
- namestr
The name of the evaluator, set to ‘hyperparameter_tuning’
- primary_colorstr
Primary color for plot elements
- secondary_colorstr
Secondary color for plot elements
- accent_colorstr
Accent color for plot elements
- categorical_columnsList[str]
List of categorical parameter columns for parallel coordinate plots
- evaluate(model: BaseEstimator, method: str, X_train: DataFrame, y_train: Series, scorer: str, kf: int, num_rep: int, n_jobs: int, plot_results: bool = False, filename: str = 'hyperparameter_tuning') BaseEstimator[source]#
Perform hyperparameter tuning using grid or random search.
Executes hyperparameter optimization using the specified search method and returns the best model found. Optionally generates performance visualizations for the tuning results.
- Parameters:
- modelbase.BaseEstimator
The model to be tuned
- methodstr
The search method to use (“grid” or “random”)
- X_trainpd.DataFrame
The training features
- y_trainpd.Series
The training target values
- scorerstr
The scoring metric to optimize
- kfint
Number of folds for cross-validation
- num_repint
Number of repetitions for cross-validation
- n_jobsint
Number of parallel jobs to run
- plot_resultsbool, optional
Whether to generate performance plots, by default False
- filenamestr, optional
Filename for saving plots, by default “hyperparameter_tuning”
- Returns:
- base.BaseEstimator
The tuned model with optimal hyperparameters