Skip to main content
. 2022 Mar 16;10(3):686. doi: 10.3390/biomedicines10030686

Table 4.

Kruskal–Wallis analysis between MRMR and Chi-square feature selection methods, respectively, using 2 and 4 classes of outcome.

χ2 p ε2
2 classes Accuracy 0.461 0.794 0.0271
Precision 0.467 0.792 0.0275
Precision M 1.207 0.547 0.071
Recall 0.467 0.792 0.0275
Recall M 0.457 0.796 0.0269
F1-Score 0.467 0.792 0.0275
F1-Score M 0.668 0.716 0.0393
4 classes Accuracy 3.27 0.195 0.1921
Precision 3.28 0.194 0.1929
Precision M 1.3 0.523 0.0764
Recall 3.28 0.194 0.1929
Recall M 2.54 0.281 0.1493
F1-Score 3.28 0.194 0.1929
F1-Score M 2.14 0.343 0.1259