Table 3.

The values of various metrics for each method on the training and test sets.

		Train	Test
Group	Model	AUC	AUC	Accuracy		Precision	Recall
T1	Lasso+SVM	0.936	0.933	0.832		0.851	0.830
	Lasso+LR	0.940	0.935	0.817		0.840	0.824
	LR1+SVM	0.952	0.951	0.855		0.869	0.858
	LR1	0.951	0.947	0.848		0.861	0.857
T1C	Lasso+SVM	0.955	0.953	0.870		0.901	0.848
	Lasso+LR	0.950	0.947	0.838		0.886	0.818
	LR1+SVM	0.961	0.959	0.864		0.922	0.818
	LR1	0.958	0.954	0.858		0.919	0.799
T2	Lasso+SVM	0.954	0.953	0.856		0.873	0.854
	Lasso+LR	0.947	0.942	0.803		0.828	0.854
	LR1+SVM	0.973	0.969	0.876		0.828	0.858
	LR1	0.968	0.967	0.860		0.892	0.843
Multi-task	Lasso1,1+SVM	0.984	0.981	0.900		0.918	0.893
	Lasso1,1+LR	0.893	0.981	0.870		0.903	0.853
	Lasso2,1+SVM	0.981	0.980	0.886		0.941	0.840
	Lasso2,1+LR	0.975	0.973	0.868		0.913	0.847
	Ours+SVM	0.993	0.992	0.920		0.969	0.847
	Ours	0.987	0.984	0.895		0.954	0.838

Bold values indicate the largest metrics indifferent models in the same sequence task.