Skip to main content
. 2019 May 23;19(10):2370. doi: 10.3390/s19102370

Figure 8.

Figure 8

The classification performance of DFDL and SVM_rbf for four essential hand movements using the targeted method. The performance of DFDL was superior to SVM_rbf in every number of channel subsets. Moreover, a decrease in performance was not observed until the number of channels was reduced to 2 in DFDL. Each dot indicates the classification accuracy of each subject. The error bar represents the standard deviation. Bold values indicate significant differences.