ROC Curve & AUC Calculator
Analyze binary classification performance with ROC curves and confusion matrices
Configuration
Number of Samples:
Class Separation:
High (Good Model)
Medium (Fair Model)
Low (Poor Model)
Random (No Signal)
Generate Data & Calculate ROC
ROC Curve Analysis
Download Data
ROC Curve
Confusion Matrix
Classification Threshold
Threshold:
0.50
Performance Metrics
Threshold Analysis
Help
ROC Curve Basics
ROC:
Receiver Operating Characteristic curve
TPR:
True Positive Rate (Sensitivity/Recall)
FPR:
False Positive Rate (1 - Specificity)
AUC:
Area Under Curve - overall performance metric (0.5 to 1.0)
Performance Metrics
Sensitivity (TPR):
TP / (TP + FN) - ability to find positives
Specificity (TNR):
TN / (TN + FP) - ability to find negatives
Precision (PPV):
TP / (TP + FP) - accuracy of positive predictions
NPV:
TN / (TN + FN) - accuracy of negative predictions
Accuracy:
(TP + TN) / Total - overall correctness