Table 4.

The performance in independent tests

Test data	Sen	Spe	PPV	ACC	BA	MCC
10%	86.67%	99.74%	68.42%	99.66%	93.20%	0.77
20%	88.57%	99.68%	72.09%	99.58%	94.13%	0.80
30%	79.83%	99.75%	77.50%	99.54%	89.79%	0.78

Sen means sensitivity, Spe means specificity, PPV means positive predictive value, ACC means accuracy, BA means balanced accuracy and MCC means Matthew's correlation coefficient. The Test data column is the percentage of data that has been randomly selected as the test set. The remaining data are used as the training set.