Improved Transductive Support Vector Machine for a Small Labelled Set in Motor Imagery-Based Brain-Computer Interface

Figure 4

Average classification accuracy (%), with varying numbers of labelled trials (M) and ratios of positive to negative labelled trials (R) with the two datasets: (a) dataset IV-a (R = 1 : 4); (b) dataset II-a (R = 1 : 4); (c) dataset IV-a (R = 2 : 3); (d) dataset II-a (R = 2 : 3); (e) dataset IV-a (R = 3 : 2); (f) dataset II-a (R = 3 : 2); (g) dataset IV-a (R = 4 : 1); (h) dataset II-a (R = 4 : 1).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)