Table 3.

ACC & AUC versus percentage of different imputation methods using Dataset1, where the training set contains 27 patients with seven time points and testing set consists of the remaining 26 patients with less than seven time points.

Metrics	Imputation method	REP-SVM	EN-LR	SVM
ACC	Tensor completion	0.790	0.751	0.775
	Median	0.724	0.713	0.721
	Mean	0.736	0.706	0.722
	KNN	0.738	0.723	0.721
AUC	Tensor completion	0.884	0.805	0.853
	Median	0.838	0.743	0.812
	Mean	0.835	0.739	0.812
	KNN	0.837	0.748	0.798

Percentage of missing values in the training set 0.23% and that in the testing set is 18.1%.