Skip to main content

Table 2 Performance of machine learning approaches on depression training dataset

From: Machine learning algorithms for systematic review: reducing workload in a preclinical review of animal studies and reducing human screening error

 

Approach 1

Approach 2

Training set size

5749

5749

Optimal cut-off score

0.1

0.07

Sensitivity

98.7%

98.7%

Upper 95% CI

0.997

0.997

Lower 95% CI

0.949

0.949

Specificity

86.0%

84.7%

Precision

50%

47.66%

Accuracy

1096/1251 = 87.6%

1081/1251 = 86.4%

WSS@95%

0.705

0.693

LR+

7.421

9.451