Assessment of eye fatigue caused by head-mounted displays using eye-tracking

Table 5 Accuracy of the three kinds of classifications

Feature	Accuracy of two-class classification				Accuracy of three-class classification				Accuracy of four-class classification
Feature	Kernel 1	Kernel 2	Kernel 3	Kernel 4	Kernel 1	Kernel 2	Kernel 3	Kernel 4	Kernel 1	Kernel 2	Kernel 3	Kernel 4
Feature set 1	0.8323	0.8310	0.8419	0.8503	0.6953	0.6849	0.6855	0.6481	0.6527	0.6269	0.6249	0.6163
Feature set 2	0.8353	0.8358	0.8460	0.8518	0.7042	0.6936	0.6955	0.6585	0.6565	0.6572	0.6530	0.6356
Feature set 3	0.8406	0.8404	0.8511	0.8542	0.7134	0.7040	0.7053	0.6709	0.6639	0.6886	0.6659	0.6454
Feature set 4	0.8511	0.8498	0.8608	0.8604	0.7308	0.7217	0.7252	0.6956	0.6778	0.7015	0.6844	0.6647
Feature set 5	0.8556	0.8550	0.8678	0.8637	0.7401	0.7319	0.7355	0.6964	0.6841	0.7084	0.6938	0.6743
Feature set 6	0.8603	0.8608	0.8751	0.8685	0.7496	0.7394	0.7450	0.7076	0.6913	0.7169	0.7037	0.6751
Feature set 7	0.8649	0.8644	0.8805	0.8720	0.7589	0.7489	0.7542	0.7190	0.6989	0.7209	0.7044	0.6846
Feature set 8	0.8708	0.8695	0.8869	0.8755	0.7690	0.7575	0.7651	0.7315	0.7053	0.7394	0.7145	0.6932
Feature set 9	0.8754	0.8742	0.8934	0.8781	0.7885	0.7656	0.7775	0.7435	0.7114	0.7398	0.7239	0.7040
Feature set 10	0.8887	0.8780	0.9079	0.8836	0.7947	0.7750	0.7853	0.7535	0.7276	0.7425	0.7248	0.7137
Blink set	0.8632	0.8479	0.8763	0.8613	0.7458	0.7267	0.7416	0.6952	0.6874	0.7225	0.7104	0.6739
Mean	0.8580	0.8552	0.8716	0.8654	0.7446	0.7317	0.7378	0.7018	0.6870	0.7059	0.6912	0.6710

ISSN: 1475-925X