Skip to main content

Table 4 AUC-ROC of COVID-19 mortality prediction results by using T-LSTM on different sets at different timestamps

From: Predicting COVID-19 disease progression and patient outcomes based on temporal deep learning

  Trainingb Validationb Testb
0 days earlya 0.996 ± 0.01 0.997 ± 0.01 0.997 ± 0.00
3 days early 0.989 ± 0.00 0.987 ± 0.01 0.969 ± 0.01
6 days early 0.960 ±0.01 0.957 ±0.0 2 0.947 ±0.03
9 days early 0.944 ±0.00 0.935 ± 0.01 0.921 ±0.03
12 days early 0.926 ± 0.01 0.924 ±0.0 2 0.914 ±0.02
15 days early 0.891 ± 0.01 0.883 ± 0.01 0.863 ±0.01
18 days early 0.852 ± 0.01 0.834 ±0.0 2 0.819 ±0.0 2
  1. an days early: The model makes a prediction n days before the final death/survival time. It uses sequence data from day 0 to n days before the last time to predict
  2. bWe use the records of 375 patients as the training set; the ratio of training set to verification set is 0.8:0.2. The records of 110 patients make up the test set. This experiment is conducted on 5-fold cross-validation