ROC-AUC Metric

What is ROC-AUC?

ROC-AUC (Receiver Operating Characteristic - Area Under Curve) measures a classifier's ability to distinguish between classes across all possible classification thresholds.

Components

ROC Curve

• Plots True Positive Rate (Recall) vs. False Positive Rate
• Shows trade-off between sensitivity and specificity

AUC (Area Under Curve)

• Single number summarizing the ROC curve
• Range: 0.5 (random) to 1.0 (perfect)

Interpretation

• 0.5: No better than random guessing
• 0.7-0.8: Acceptable discrimination
• 0.8-0.9: Excellent discrimination
• >0.9: Outstanding discrimination

When to Use

• Binary classification problems
• When you want threshold-independent evaluation
• Comparing multiple classifiers

Related Terms

• Precision: Positive predictive value
• Recall: True positive rate
• F1 Score: Harmonic mean of precision/recall