PR AUC refers to the Area Under the Precision-Recall Curve. It is a single scalar metric that summarizes a binary classifier’s performance across different threshold settings by measuring the area under the curve plotted with precision on the y-axis and recall on the x-axis​.The precision-recall (PR) curve itself shows the trade-off between precision (how many of the positive predictions are correct) and recall (how many of the actual positives are identified) as the decision threshold of the classifier is varied. A model with perfect predictions would achieve a PR AUC of 1.0, indicating it can attain 100% precision and 100% recall simultaneously at some threshold​.In contrast to the more commonly known ROC AUC, PR AUC is often preferred for imbalanced datasets or scenarios where positive class is rare, because it focuses only on performance with respect to the positive class (it does not take true negatives into account)​.A high PR AUC means that the classifier is able to maintain high precision and high recall, which is desirable in tasks like anomaly detection or medical diagnosis where one wants to catch as many positives as possible while keeping false alarms low. In summary, PR AUC is a useful evaluation metric that condenses the precision-recall curve into a single number, facilitating model comparison especially in contexts where precision and recall are more informative than overall accuracy.