Table 10.

Cross-dataset prediction accuracy results (Trained on the SEED/DEAP and tested on the LUMED).

Users (LUMED)	Trained on SEED		Trained on DEAP
	Accuracy (Pos-Neg)	Accuracy (with Median Filtering) (Pos-Neg)	Accuracy (Pos-Neg)	Accuracy (with Median Filtering) (Pos-Neg)
User 1	68.2	72.3	42.7	48.3
User 2	54.5	61.7	50.4	56.8
User 3	54.3	59.6	49.7	54.1
User 4	59.6	64.1	51.3	57.9
User 5	44.8	53.7	87.1	89.7
User 6	67.1	73.5	52.6	58.4
User 7	53.2	60.8	53.8	59.2
User 8	64.5	71.2	46.0	49.3
User 9	48.6	50.9	84.7	85.6
User 10	64.9	76.3	51.5	58.8
User 11	57.1	64.8	63.8	67.1
Average	57.89	64.44	57.6	62.29
Std. Dev.	7.33	7.82	14.23	12.91