Table 5.
Average misclassification error rate and area under the receiver operating characteristic curve (AUC) of sample classification for three class real dataset (MDA-MB-231) with 15.81% missing values and artificially imputed different rates of outliers.
| Methods | Without outliers MER (AUC) | 3% outliers MER (AUC) | 5% outliers MER (AUC) | 7% outliers MER (AUC) | 10% outliers MER (AUC) |
|---|---|---|---|---|---|
| RF | 4.23 (0.9635) | 21.57 (0.7996) | 23.53 (0.7771) | 31.96 (0.6951) | 34.70 (0.6777) |
| PPCA | 4.33 (0.9629) | 18.27 (0.8288) | 25.13 (0.7564) | 21.83 (0.7938) | 30.63 (0.7238) |
| kNN | 3.43 (0.9759) | 18.63 (0.8296) | 21.80 (0.7940) | 24.56 (0.7646) | 43.56 (0.6672) |
| BPCA | 8.07 (0.9267) | 18.77 (0.8207) | 22.06 (0.7836) | 26.43 (0.7475) | 33.70 (0.6858) |
| EM-PCA | 4.46 (0.9619) | 19.76 (0.8101) | 19.63 (0.8161) | 21.66 (0.7938) | 34.93 (0.6741) |
| Zero | 3.45 (0.9755) | 12.20 (0.8843) | 25.86 (0.7554) | 27.40 (0.7347) | 38.53 (0.6374) |
| Mean | 3.73 (0.9728) | 11.73 (0.8858) | 25.30 (0.7597) | 26.2 (0.7452) | 35.90 (0.6668) |
| Median | 3.67 (0.9732) | 11.36 (0.8861) | 22.73 (0.7874) | 23.16 (0.7753) | 34.90 (0.6748) |
| Minimum | 3.43 (0.9758) | 13.10 (0.8774) | 24.64 (0.7647) | 27.16 (0.7457) | 37.10 (0.6457) |
| rmiMAE | 1.45 (0.9854) | 2.47 (0.9757) | 3.04 (0.9753) | 3.56 (0.9668) | 4.01 (0.9611) |
| Proposed | 0.17 (0.9992) | 0.25 (0.9983) | 0.30 (0.9975) | 0.52 (0.9951) | 1.13 (0.9892) |
Bold indicates the lower MER and Higher AUC throughout the column.