quick_anomaly_detector.data_process.calculate_metrics
- quick_anomaly_detector.data_process.calculate_metrics(actual_labels, predicted_labels)
Calculate various evaluation metrics for binary classification.
- Parameters:
actual_labels (array-like) – Ground truth labels.
predicted_labels (array-like) – Predicted labels.
- Returns:
Dictionary containing the following evaluation metrics: - precision (float): Precision score. - recall (float): Recall score. - label_pass_rate (float): Proportion of samples labeled as negative class in the ground truth. - predict_pass_rate (float): Proportion of samples predicted as negative class. - ks (float): Kolmogorov-Smirnov statistic. - gini (float): Gini coefficient. - f1 (float): F1 score. - auc_roc (float): Area under the ROC curve. - accuracy (float): Accuracy score.
- Return type: