Table 1:

The results of mortality predictions

Model	MCC	F1 score	AUROC
Logreg + count	0.3664±0.01 35	0.4604±0.02 51	0.7884±0.02 28
Logreg + tf-idf	0.3509±0.02 03	0.4457±0.03 30	0.7979±0.00 94
SVM + count	0.3234±0.02 41	0.4157±0.06 79	0.7726±0.02 67
SVM + tf-idf	0.3549±0.02 18	0.4404±0.02 27	0.7899±0.01 18
GRNN + word2vec	0.4396±0.04 86	0.5074±0.06 13	0.8650±0.01 00
GRNN-HA + word2vec	0.4835±0.05 85	0.5766±0.06 23	0.8603±0.01 67