Table 3.

Comparison of the classification accuracy (in %) with other machine learning techniques based on the dataset 2b obtained from the BCI competition IV.

Subject	Proposed Method	LDA	SVM	Yousef and Ugur [13]
1	78.81%	64.58%	74.65%	76.00%
2	80.14%	58.08%	75.36%	65.80%
3	77.08%	61.11%	70.13%	75.30%
4	91.37%	67.34%	81.98%	95.30%
5	86.15%	66.55%	79.10%	83.00%
6	80.73%	59.72%	72.91%	79.50%
7	81.59%	69.79%	80.55%	74.50%
8	81.25%	67.10%	76.64%	75.30%
9	84.37%	60.06%	77.08%	73.30%
Average	82.39%	63.81%	76.49%	77.56%