Table 4.
Summary statistics for both naive logistic regression model (left) and anomaly detection model (right).
| Naive logistic regression statistics | Anomaly detection statistics | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| BIDMC | Bangalore | Bhopal | Total | BIDMC | Bangalore | Bhopal | Total | ||
| RMSE | 0.304 | 0.224 | 0.284 | 0.271 | RMSE | 0.268 | 0.180 | 0.125 | 0.191 |
| Sensitivity | 0.000 | 0.200 | 0.000 | 0.048 | Sensitivity | 0.009 | 0.004 | 0.004 | 0.006 |
| Specificity | 0.990 | 0.967 | 0.933 | 0.968 | Specificity | 0.996 | 0.997 | 0.999 | 0.997 |
| TP | 0 | 1 | 0 | 1 | TP | 9 | 3 | 1 | 13 |
| FP | 1 | 7 | 9 | 14 | FP | 54 | 97 | 24 | 175 |
| TN | 98 | 208 | 125 | 427 | TN | 13314 | 27698 | 18540 | 59552 |
| FN | 9 | 4 | 2 | 20 | FN | 978 | 834 | 269 | 2081 |
RMSE, TP, FP, TN, and FN stand for root mean squared error, true positive, false positive, true negative, and false negative respectively.