Area Under the Curve (AUC)

Area Under the Curve (AUC) is a metric used to evaluate how well a model, like a diagnostic test, can distinguish between different outcomes, such as disease vs. no disease. Imagine plotting the model’s ability to correctly identify positives versus false positives at various thresholds; the AUC measures the total area under this plot. A higher AUC indicates better discrimination, with 1.0 representing perfect accuracy and 0.5 suggesting no better than random chance. Essentially, AUC summarizes the overall performance of a model across all possible decision points in a single, interpretable number.