slim_gsgp.evaluators

slim_gsgp.evaluators.fitness_functions

This module provides various error metrics functions for evaluating machine learning models.

slim_gsgp.evaluators.fitness_functions.mae(y_true: Tensor, y_pred: Tensor) → Tensor[source]

Compute Mean Absolute Error (MAE).

Parameters:

y_true (torch.Tensor) – True values.
y_pred (torch.Tensor) – Predicted values.

Returns:

MAE value.

Return type:

torch.Tensor

slim_gsgp.evaluators.fitness_functions.mae_int(y_true: Tensor, y_pred: Tensor) → Tensor[source]

Compute Mean Absolute Error (MAE) for integer values.

Parameters:

y_true (torch.Tensor) – True values.
y_pred (torch.Tensor) – Predicted values.

Returns:

MAE value for integer predictions.

Return type:

torch.Tensor

slim_gsgp.evaluators.fitness_functions.mse(y_true: Tensor, y_pred: Tensor) → Tensor[source]

Compute Mean Squared Error (MSE).

Parameters:

y_true (torch.Tensor) – True values.
y_pred (torch.Tensor) – Predicted values.

Returns:

MSE value.

Return type:

torch.Tensor

slim_gsgp.evaluators.fitness_functions.r2_score(y_true: Tensor, y_pred: Tensor) → Tensor[source]

Compute R-squared (R²) score.

If using this fitness function, please ensure that you are maximizing the fitness value when

Parameters:

y_true (torch.Tensor) – True values.
y_pred (torch.Tensor) – Predicted values.

Returns:

R² score value.

Return type:

torch.Tensor

slim_gsgp.evaluators.fitness_functions.rmse(y_true: Tensor, y_pred: Tensor) → Tensor[source]

Compute Root Mean Squared Error (RMSE).

Parameters:

y_true (torch.Tensor) – True values.
y_pred (torch.Tensor) – Predicted values.

Returns:

RMSE value.

Return type:

torch.Tensor

slim_gsgp.evaluators.fitness_functions.signed_errors(y_true: Tensor, y_pred: Tensor) → Tensor[source]

Compute signed errors between true and predicted values.

Parameters:

y_true (torch.Tensor) – True values.
y_pred (torch.Tensor) – Predicted values.

Returns:

Signed error values.

Return type:

torch.Tensor