Table 5.
Comparison of the performance in RUTI prediction models after hospitalization for UTI through fivefold cross validation (sample size = 809). Abbreviations: UTI, urinary tract infection; RUTI, recurrent urinary tract infection.
| Algorithm | Accuracy | Sensitivity | Specificity | |||
|---|---|---|---|---|---|---|
| Mean | Standard deviation | Mean | Standard deviation | Mean | Standard deviation | |
| Logistic regression | 0.604 | 0.026 | 0.590 | 0.065 | 0.605 | 0.034 |
| Decision tree | 0.635 | 0.052 | 0.600 | 0.061 | 0.640 | 0.057 |
| Random forest | 0.709 | 0.047 | 0.620 | 0.057 | 0.722 | 0.058 |