AUC-ROC
Area Under the Receiver Operating Characteristic Curve -- a metric that measures a classification model's ability to distinguish between classes across all decision thresholds.
What It Measures
The ROC curve plots True Positive Rate vs. False Positive Rate at every possible classification threshold. AUC is the area under this curve. AUC = 1.0 means perfect classification. AUC = 0.5 means random guessing.
Why It's Useful
Unlike accuracy, AUC is threshold-independent -- it evaluates the model's ranking ability across all thresholds. It's also robust to class imbalance, making it preferred for problems like fraud detection or disease diagnosis.
Interpretation
AUC = 0.9 means: if you randomly pick one positive and one negative example, the model gives the positive a higher score 90% of the time.