Skip to main content
. 2021 Jul 30;33(24):17621–17632. doi: 10.1007/s00521-021-06346-3

Table 4.

Effect of the SFS method on classification results

Classifier Used Features ACC
(after used SFS)
ACC
(before used SFS)
Polynomial-SVM 19 0.9556 0.9556
RBF-SVM 18 0.9667 0.9611
Linear-LDA 11 0.8388 0.8111
Quadratic-LDA 15 0.9111 0.9056
Euclidean-kNN 19 0.9833 0.9833
Chebychev-kNN 13 0.9444 0.9056
PLSR 17 0.8277 0.8111

The highest value in each classifier is indicated in boldface