Skip to main content
. 2021 Jun 22;15:656578. doi: 10.3389/fnhum.2021.656578

Table 2.

The highest classification accuracy using the ordinary linear support vector machine (SVM) and the l1-regularized SVM, respectively.

Methods Stimulus λ No. of selected features Accuracy (%)
Ordinal linear SVM Positive 81.3
Negative 78.1
l1-regurarized SVM Positive 0.13 10 93.8
Negative 0.24 21 87.5