ETIA.AFS.predictive_model

class PredictiveModel[source]

Bases: object

A class for creating and training predictive models.

random_forest(config, target_type)[source]

Creates a Random Forest model based on the configuration and target type.

linear_regression()[source]

Creates a Linear Regression model.

fit(config, train_X, train_y, selected_features, preprocessor, target_type)[source]

Fits the model to the training data using the specified configuration.

predict(X)[source]

Makes predictions using the trained model.

random_forest(config: Dict[str, Any], target_type: str)[source]

Creates a Random Forest model based on the configuration and target type.

Parameters:

  • config (dict) – Configuration settings for the Random Forest model, including hyperparameters like n_estimators, min_samples_leaf, and max_features.

  • target_type (str) – The type of the target variable (‘categorical’ for classification, ‘continuous’ for regression).

Returns:

model – The initialized Random Forest model.

Return type:

RandomForestClassifier or RandomForestRegressor

linear_regression()[source]

Creates a Linear Regression model.

Returns:

model – The initialized Linear Regression model.

Return type:

LinearRegression

fit(config: Dict[str, Any], train_X: Any, train_y: Any, selected_features: Any, preprocessor: Any | None, target_type: str)[source]

Fits the model to the training data.

Parameters:

  • config (dict) – Configuration settings for the model, including the type of model (‘random_forest’ or ‘linear_regression’).

  • train_X (array-like) – Training data for the input variables.

  • train_y (array-like) – Training data for the target variable.

  • selected_features (any) – The features selected for model training.

  • preprocessor (object, optional) – A preprocessor object that can be used to transform the input data. Default is None.

  • target_type (str) – The type of the target variable (‘categorical’ or ‘continuous’).

Raises:

ValueError – If an unsupported model type is specified in the configuration.

predict(X: Any) → Any[source]

Makes predictions using the trained model.

Parameters:

X (array-like) – The input data for which predictions are to be made.

Returns:

predictions – The predicted values based on the input data.

Return type:

array-like