ml_grid.pipeline.grid_search_cross_validate

Classes

grid_search_crossvalidate

Initializes and runs a cross-validated hyperparameter search.

Functions

`dummy_auc`(→ float)	Returns a constant AUC score of 0.5.
`scale_data`(→ pandas.DataFrame)	Scales the data to a [0, 1] range if it's not already scaled.

Module Contents

class ml_grid.pipeline.grid_search_cross_validate.grid_search_crossvalidate(algorithm_implementation: Any, parameter_space: Dict | List[Dict], method_name: str, ml_grid_object: Any, sub_sample_parameter_val: int = 100)[source]

Initializes and runs a cross-validated hyperparameter search.

This class takes a given algorithm, its parameter space, and data from the main pipeline object to perform either a grid search, randomized search, or Bayesian search for the best hyperparameters. It then logs the results.

Parameters:

algorithm_implementation (Any) – The scikit-learn compatible estimator instance.
parameter_space (Union[Dict, List[Dict]]) – The dictionary or list of dictionaries defining the hyperparameter search space.
method_name (str) – The name of the algorithm method.
ml_grid_object (Any) – The main pipeline object containing all data (X_train, y_train, etc.) and parameters for the current iteration.
sub_sample_parameter_val (int, optional) – A value used to limit the number of iterations in a randomized search. Defaults to 100.