ROC-AUC
A metric measuring a classifier's ability to distinguish between classes across all thresholds.
Overview
ROC-AUC (Receiver Operating Characteristic - Area Under the Curve) measures a binary classifier's ability to distinguish between positive and negative classes across all possible classification thresholds. The ROC curve plots True Positive Rate vs False Positive Rate at each threshold, and AUC is the area under this curve.
Key Details
An AUC of 1.0 indicates perfect classification, 0.5 indicates random guessing, and below 0.5 indicates worse than random. AUC is threshold-independent, making it useful when the optimal classification threshold isn't known in advance. It's widely used in medical diagnostics, fraud detection, and any binary classification task. For multi-class problems, one-vs-rest AUC extensions are used.