Table 2.
Average misclassification error rate(MER) and area under the receiver operating characteristic curve (AUC) for two class simulated data with 5% missing values and different rates of outliers.
| Methods | 3% Outliers MER (AUC) | 5% Outliers MER (AUC) | 7% Outliers MER (AUC) | 10% Outliers MER (AUC) |
|---|---|---|---|---|
| RF | 4.10 (0.9516) | 8.67 (0.9143) | 16.17 (0.8395) | 17 (0.8332) |
| PPCA | 5.67 (0.9408) | 7.37 (0.9276) | 15.53 (0.8454) | 19.37 (0.8048) |
| kNN | 5.53 (0.9412) | 9.27 (0.9054) | 16.57 (0.83885) | 19.6 (0.8042) |
| BPCA | 5.77 (0.9391) | 8.50 (0.9149) | 16.27 (0.84025) | 21.2 (0.7882) |
| EM-PCA | 5.03 (0.9460) | 7.43 (0.9270) | 15.33 (0.8475) | 15.77 (0.8385) |
| Zero | 7.73 (0.9224) | 12.70 (0.8759) | 19 (0.8068) | 18.9 (0.8114) |
| Mean | 5.93 (0.9371) | 9.30 (0.9059) | 16.5 (0.8358) | 21.37 (0.7858) |
| Median | 6.17 (0.9353) | 9.67 (0.9021) | 16.43 (0.8366) | 19.87 (0.8021) |
| Minimum | 8.67 (0.9088) | 13.70 (0.8675) | 18.87 (0.8088) | 19.27 (0.8073) |
| rmiMAE | 1.69 (0.9831) | 1.81 (0.9819) | 2.36 (0.9764) | 2.82 ( 0.9718) |
| Proposed | 1.27 (0.9882) | 1.30 (0.9878) | 1.30 (0.9876) | 1.32 (0.9872) |
Bold indicates the lower MER and Higher AUC throughout the column.