Skip to main content
. 2021 May 29;134:104531. doi: 10.1016/j.compbiomed.2021.104531

Table 3.

Performance comparison of the machine learning models for COVID-19.

Metric Diagnostic model
Disease severity model
RNA DT PLS-DA K-NN RNA DT PLS-DA K-NN
Training time 21 min. 43 s 27 min. 11 s 31 min. 19 s 22 min. 15 s 7 min. 1 s 10 min. 19 s 18 min. 3 s 09 min. 53 s
Calibration error 1.0% 0.5% 1.2% 0.5% 1.0% 8.4% 6.0% 0.4%
Cross validation error 0.8% 1.0% 0.9% 0.6% 0.5% 1.8% 4.0% 0.7%
Sensibility 0.93 0.89 0.88 0.84 0.99 0.90 0.87 0.82
Specificity 0.94 0.89 0.90 0.83 0.97 0.94 0.88 0.88
Accuracya 0.94 0.90 0.90 0.84 0.98 0.92 0.88 0.86
a

Area under the ROC curve.