Feature Selection in Classification of Eye Movements Using Electrooculography for Activity Recognition

Table 3

Detailed performance of the proposed approach with CBFS and mRMR features.

Activity	Features	TP (%)	FP (%)	Precision (%)	Recall (%)	-Measure (%)
Null	DEFS	75.31	14.20	84.14	92.24	87.99
	mRMR	65.50	15.41	80.95	93.32	86.70
	CBFS	67.60	14.30	82.54	89.89	86.06
Read	DEFS	71.60	17.12	80.70	93.74	86.73
	mRMR	58.30	18.29	76.11	91.79	83.22
	CBFS	57.29	16.58	77.56	82.27	79.84
Browse	DEFS	75.17	14.15	84.16	93.75	88.69
	mRMR	68.98	14.38	82.75	95.17	88.53
	CBFS	54.00	16.50	76.60	81.82	79.12
Write	DEFS	81.23	09.92	89.12	97.05	92.91
	mRMR	79.10	08.85	88.94	94.56	92.19
	CBFS	69.24	12.10	85.12	90.39	87.68
Video	DEFS	85.51	10.20	89.34	99.38	94.09
	mRMR	67.41	13.09	83.74	90.23	86.86
	CBFS	62.37	15.44	80.16	86.16	83.05
Copy	DEFS	80.21	13.18	85.89	97.61	91.37
	mRMR	70.34	10.28	85.25	87.66	87.45
	CBFS	56.44	11.17	83.48	81.88	82.67