Data & Analytics
AUC (Area Under the Curve)
Also known as:
AUC
AUC-ROC
AUROC
Area Under ROC
Updated: 2/12/2026
The area under the ROC curve – a single number (0-1) summarizing the overall quality of a binary classifier.
Quick Summary
AUC summarizes the ROC curve in one number – the standard metric for binary classification.
Explanation
AUC = 0.5 equals random guessing, AUC = 1.0 perfect separation. Interpretable as probability that a positive example ranks higher than a negative one.
Marketing Relevance
AUC is the most widely used metric for model comparison in Kaggle, research, and industry.
Common Pitfalls
AUC hides the optimal threshold. With severe imbalance, high AUC can still mean poor precision.
Origin & History
AUC was derived from signal detection theory (1960s) and has been the dominant ML classification metric since the 2000s.
Comparisons & Differences
AUC (Area Under the Curve) vs. Log Loss
AUC measures ranking quality; Log Loss measures calibration quality.