Table 1.

Performance of the machine learning classifiers using binary profile on balanced data set for various window sizes

Pattern (classifier)	Training data set					Validation data set
	Sen	Spc	Acc	MCC	AUROC	Sen	Spc	Acc	MCC	AUROC
Pat5(SVC)	67.42	60.90	64.16	0.28	0.71	67.18	55.69	61.43	0.23	0.68
Pat7(SVC)	66.52	64.12	65.32	0.31	0.72	66.26	60.00	63.13	0.26	0.70
Pat9(RF)	65.39	65.77	65.58	0.31	0.73	65.69	59.69	62.69	0.25	0.70
Pat11(RF)	66.07	65.17	65.62	0.31	0.72	69.08	60.62	64.85	0.30	0.71
Pat13(RF)	65.62	65.77	65.69	0.31	0.72	69.69	62.31	66.00	0.32	0.71
Pat15(RF)	66.52	65.24	65.88	0.32	0.72	68.00	59.23	63.62	0.27	0.71
Pat17(RF)	67.64	61.12	64.38	0.29	0.71	68.15	58.92	63.54	0.27	0.69
Pat19(RF)	66.37	62.47	64.42	0.29	0.71	67.69	59.54	63.62	0.27	0.70
Pat21(RF)	67.87	61.57	64.72	0.29	0.71	67.38	60.15	63.77	0.28	0.70
Pat23(RF)	67.57	62.02	64.79	0.30	0.71	66.00	59.85	62.92	0.26	0.69

Note: Various classifiers were used for building models and the performance obtained by the best classifier (mentioned in the bracket) for each window size has been reported.