Comparison of performance with the existing automated candidate classifiers on HTRU-Medlat. Compared with these existing results on HTRU-S, the performance was improved with fewer features.
| . | Algor. . | Feature Num . | Sampling . | Classifier . | Features ID . | Recall (per cent) . | FPR (per cent) . |
|---|---|---|---|---|---|---|---|
| Our Work | GS | 2 | SMOTE | Adaboost | 1,10 | 99 | 0.651 |
| RFE | 3 | SMOTE | GBoost | 3,5,10 | 99 | 0.162 | |
| RFE | 4 | SMOTE | GBoost | 3,5,10,11 | 99 | 0.102 | |
| Bates et al. (2012) | / | 22 | NonSampling | ANN | / | 85 | 1 |
| Morello et al. (2014) | / | 6 | NonSampling | SPINN | 1,2,3,4,5,6 | 98/99/100 | 0.01/0.11/0.64 |
| Lyon et al. (2016) | / | 8 | Data Stream | GHVFDT | 7,8,9,10,11,12,13,14 | 92.8 | 0.5 |
| . | Algor. . | Feature Num . | Sampling . | Classifier . | Features ID . | Recall (per cent) . | FPR (per cent) . |
|---|---|---|---|---|---|---|---|
| Our Work | GS | 2 | SMOTE | Adaboost | 1,10 | 99 | 0.651 |
| RFE | 3 | SMOTE | GBoost | 3,5,10 | 99 | 0.162 | |
| RFE | 4 | SMOTE | GBoost | 3,5,10,11 | 99 | 0.102 | |
| Bates et al. (2012) | / | 22 | NonSampling | ANN | / | 85 | 1 |
| Morello et al. (2014) | / | 6 | NonSampling | SPINN | 1,2,3,4,5,6 | 98/99/100 | 0.01/0.11/0.64 |
| Lyon et al. (2016) | / | 8 | Data Stream | GHVFDT | 7,8,9,10,11,12,13,14 | 92.8 | 0.5 |
Comparison of performance with the existing automated candidate classifiers on HTRU-Medlat. Compared with these existing results on HTRU-S, the performance was improved with fewer features.
| . | Algor. . | Feature Num . | Sampling . | Classifier . | Features ID . | Recall (per cent) . | FPR (per cent) . |
|---|---|---|---|---|---|---|---|
| Our Work | GS | 2 | SMOTE | Adaboost | 1,10 | 99 | 0.651 |
| RFE | 3 | SMOTE | GBoost | 3,5,10 | 99 | 0.162 | |
| RFE | 4 | SMOTE | GBoost | 3,5,10,11 | 99 | 0.102 | |
| Bates et al. (2012) | / | 22 | NonSampling | ANN | / | 85 | 1 |
| Morello et al. (2014) | / | 6 | NonSampling | SPINN | 1,2,3,4,5,6 | 98/99/100 | 0.01/0.11/0.64 |
| Lyon et al. (2016) | / | 8 | Data Stream | GHVFDT | 7,8,9,10,11,12,13,14 | 92.8 | 0.5 |
| . | Algor. . | Feature Num . | Sampling . | Classifier . | Features ID . | Recall (per cent) . | FPR (per cent) . |
|---|---|---|---|---|---|---|---|
| Our Work | GS | 2 | SMOTE | Adaboost | 1,10 | 99 | 0.651 |
| RFE | 3 | SMOTE | GBoost | 3,5,10 | 99 | 0.162 | |
| RFE | 4 | SMOTE | GBoost | 3,5,10,11 | 99 | 0.102 | |
| Bates et al. (2012) | / | 22 | NonSampling | ANN | / | 85 | 1 |
| Morello et al. (2014) | / | 6 | NonSampling | SPINN | 1,2,3,4,5,6 | 98/99/100 | 0.01/0.11/0.64 |
| Lyon et al. (2016) | / | 8 | Data Stream | GHVFDT | 7,8,9,10,11,12,13,14 | 92.8 | 0.5 |
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
