Skip to main content
. 2020 Oct 27;8(10):e20324. doi: 10.2196/20324

Table 2.

Areas under the receiver operating characteristic curve and 95% confidence intervals of hospital admission prediction models according to machine learning methods and prediction models.

Model type Model 1a Model 2b Model 3c Model 4d
Logistic regression 0.631
(0.602-0.657)		 0.750
(0.723-0.774)		 0.805
(0.782-0.827)		 0.750
(0.720-0.774)
Lasso 0.631
(0.602-0.657)		 0.755
(0.730-0.779)		 0.817
(0.793-0.839)		 0.811
(0.787-0.832)
Random forest 0.594
(0.567-0.619)		 0.735
(0.710-0.763)		 0.813
(0.786-0.834)		 0.814
(0.786-0.833)
Gradient boosting machine 0.624
(0.598-0.652)		 0.758
(0.734-0.783)		 0.818
(0.792-0.839)		 0.815
(0.788-0.833)

aModel 1: Age and sex.

bModel 2: Age, sex, and chief complaint.

cModel 3: Age, sex, chief complaint, and vital signs.

dModel 4: Age, sex, chief complaint, vital signs, and past medical history.