Source code for caliber.binary_classification.minimizing.linear_scaling.performance.bal_acc_linear_scaling
from typing import Optional
import numpy as np
from caliber.binary_classification.minimizing.linear_scaling.performance.base import (
PerformanceLinearScalingBinaryClassificationModel,
)
[docs]
class BalancedAccuracyLinearScalingBinaryClassificationModel(
PerformanceLinearScalingBinaryClassificationModel
):
def __init__(
self,
threshold: float,
lam: float = 0.01,
minimize_options: Optional[dict] = None,
has_intercept: bool = True,
):
super().__init__(
loss_fn=_balanced_accuracy_loss_fn,
threshold=threshold,
minimize_options=minimize_options,
has_intercept=has_intercept,
)
self._lam = lam
def _balanced_accuracy_loss_fn(targets: np.ndarray, preds: np.ndarray) -> float:
n_pos_targets = np.sum(targets)
n_neg_targets = len(targets) - n_pos_targets
tpr = np.sum(targets * preds) / n_pos_targets if n_pos_targets > 0 else 0.0
tnr = (
np.sum((1 - targets) * (1 - preds)) / n_neg_targets
if n_neg_targets > 0
else 0.0
)
return -(tpr + tnr)