Table 6.
The performance metrics of the neural network model using Adam using the default threshold (50%) after matching.
| Sociodemographic groups | Recall | Specificity | Accuracy | AUCa | Differenceb | P value | |||
| Sex | 2.82 | <.001 | |||||||
|
|
Male | 57.14 | 96.80 | 96.63 | 92.65 |
|
|
||
|
|
Female | 59.26 | 96.10 | 95.79 | 92.34 |
|
|
||
| Marital status | 11.12 | <.001 | |||||||
|
|
Never been married | 64.71 | 96.74 | 96.58 | 95.75 |
|
|
||
|
|
Other groups of marital status | 54.17 | 96.16 | 95.84 | 90.33 |
|
|
||
| Working condition | 25.25 | <.001 | |||||||
|
|
Working ≥35 hours | 77.78 | 96.00 | 95.90 | 97.72 |
|
|
||
|
|
Other groups of working condition | 53.13 | 96.60 | 96.32 | 90.74 |
|
|
||
| Race | 67.63 | <.001 | |||||||
|
|
White | 65.52 | 95.90 | 95.71 | 95.72 |
|
|
||
|
|
Black | 0.00 | 98.01 | 97.52 | 76.74 |
|
|
||
| Income | 5.40 | <.001 | |||||||
|
|
An income of <US $20,000 | 60.00 | 94.14 | 93.99 | 93.36 |
|
|
||
|
|
Other groups of income | 58.06 | 97.60 | 97.32 | 92.54 |
|
|
||
aAUC: area under the curve.
bDifference between recall and specificity values.