Table 3.
Machine learning (ML) models did not accurately detect the postpartum depression (PPD)–equivalent period in women without PPD.
Time period and metric | Value, mean (SD) | ||
Prepregnancy period | |||
|
Sensitivity | 0.89 (0.15) | |
|
Specificity | 0.91 (0.10) | |
|
Precision | 0.88 (0.09) | |
|
Recall | 0.89 (0.15) | |
|
F1-score | 0.88 (0.13) | |
Pregnancy period | |||
|
Sensitivity | 0.85 (0.12) | |
|
Specificity | 0.91 (0.06) | |
|
Precision | 0.87 (0.07) | |
|
Recall | 0.85 (0.12) | |
|
F1-score | 0.86 (0.09) | |
Postpartum period | |||
|
Sensitivity | 0.74 (0.20) | |
|
Specificity | 0.96 (0.04) | |
|
Precision | 0.76 (0.16) | |
|
Recall | 0.74 (0.20) | |
|
F1-score | 0.75 (0.18) | |
PPD-equivalent period | |||
|
Sensitivity | 0.52 (0.33) | |
|
Specificity | 0.99 (0.03) | |
|
Precision | 0.69 (0.28) | |
|
Recall | 0.52 (0.33) | |
|
F1-score | 0.61 (0.30) |