Table 1.
The list of 35 algorithms to be classified and their abbreviations.
Table 2.
SDM methods based on single, double, and multiple fpfs-matrices, which pass five tests.
Table 3.
Classification of the SDM methods utilizing single fpfs-matrices.
Table 4.
Classification of the SDM methods utilizing double fpfs-matrices.
Table 5.
SDM methods utilizing multiple fpfs-matrices.
Table 6.
Description of UCI data sets.
Table 7.
Performance comparisons of algorithms on the Parkinsons[sic] dataset.
Table 8.
Performance comparisons of algorithms on the Wine dataset.
Table 9.
Performance comparisons of algorithms on the Sonar dataset.
Table 10.
Performance comparisons of algorithms on the Hayes dataset.
Table 11.
Performance comparisons of algorithms on the Libras Movement.
Table 12.
Performance comparisons of algorithms on the Teaching dataset.
Table 13.
Performance comparisons of algorithms on the Ionosphere dataset.
Table 14.
Performance comparisons of algorithms on the Whosalers dataset.
Table 15.
Performance comparisons of algorithms on the Glass dataset.
Table 16.
Performance comparisons of algorithms on the Ecoli dataset.
Table 17.
Ranking Single fpfs-matrices for ten datasets.
Table 18.
Ranking double fpfs-matrices for ten datasets.
Table 19.
Sorting more than two fpfs-matrices for ten datasets.
Fig 1.
The Nemenyi diagrams (at 0.05 significance levels) of the mean Acc performances of 35 SDM methods.
Fig 2.
The Nemenyi diagrams (at 0.05 significance levels) of the mean Pre performances of 35 SDM methods.
Fig 3.
The Nemenyi diagrams (at 0.05 significance levels) of the mean Rec performances of 35 SDM methods.
Fig 4.
The Nemenyi diagrams (at 0.05 significance levels) of the mean Spe performances of 35 SDM methods.
Fig 5.
The Nemenyi diagrams (at 0.05 significance levels) of the mean F1-score performances of 35 SDM methods.
Fig 6.
The Nemenyi diagrams (at 0.05 significance levels) of the Running Times performances of 35 SDM methods.
Table 20.
Computational complexity of the SDM methods.