Table 1. Sensitivity, specificity, and accuracy of urban blight algorithm.
Model 1 | ||||
Urban blight = 1 | Urban blight = 0 | PPV (Sensitivity) | 91% | |
Pred(Urban blight = 1) | 210,413 | 43,701 | NPV (Specificity) | 55% |
Pred(Urban blight = 0) | 20,718 | 54,211 | ACC (Accuracy) | 80% |
Model 2 | ||||
Urban blight = 1 | Urban blight = 0 | PPV (Sensitivity) | 90% | |
Pred(Urban blight = 1) | 208,947 | 42,172 | NPV (Specificity) | 57% |
Pred(Urban blight = 0) | 22,184 | 55,740 | ACC (Accuracy) | 80% |
Model 3 | ||||
Urban blight = 1 | Urban blight = 0 | PPV (Sensitivity) | 90% | |
Pred(Urban blight = 1) | 208,213 | 23,884 | NPV (Specificity) | 76% |
Pred(Urban blight = 0) | 22,918 | 74,028 | ACC (Accuracy) | 86% |