. 2020 Oct 19;10(10):e040132. doi: 10.1136/bmjopen-2020-040132

Table 3.

The number of actual and predicted outcomes of prediction models in the test set

Prediction model	Classification	Perinatal status
Prediction model	Classification	Alive	Died
	Actual number of events	12 227	468
Logistic regression	Correctly predicted outcome	10 627	261
Logistic regression	Incorrectly predicted outcome	1600	207
Artificial neural network	Correctly predicted outcome	10 225	280
Artificial neural network	Incorrectly predicted outcome	2002	188
Random Fforests	Correctly predicted outcome	10 774	251
Random Fforests	Incorrectly predicted outcome	1453	217
Naïve bayes	Correctly predicted outcome	10 386	267
Naïve bayes	Incorrectly predicted outcome	1841	201
Bagging	Correctly predicted outcome	10 175	260
Bagging	Incorrectly predicted outcome	2052	208
Boosting	Correctly predicted outcome	10 852	252
Boosting	Incorrectly predicted outcome	1375	216