Research Article
Recurrent Adaptive Classifier Ensemble for Handling Recurring Concept Drifts
Table 4
Kappa M values for all the ten datasets.
| | Dataset | RACE | CALMID | DUE | SOFEnsemble | AWDOB |
| | Hyperplane | 78.37 (1) | 69.54 (2) | 62.56 (4) | 66.48 (3) | 61.42 (5) | | Stagger | 76.48 (3) | 74.39 (2) | 67.42 (4) | 79.56 (1) | 65.42 (5) | | LED | 81.36 (1) | 69.35 (4) | 72.49 (3) | 80.48 (2) | 67.43 (5) | | SEA | 75.34 (3) | 77.49 (2) | 66.43 (5) | 83.28 (1) | 69.78 (4) | | Random Tree | 82.37 (1) | 74.36 (3) | 67.58 (5) | 69.42 (4) | 78.48 (2) | | Airlines | 80.59 (2) | 72.46 (3) | 62.34 (5) | 83.73 (1) | 64.32 (4) | | KDD99 | 66.28 (4) | 64.54 (3) | 79.36 (2) | 62.48 (5) | 84.42 (1) | | Covertype | 77.36 (1) | 73.37 (3) | 67.37 (4) | 74.46 (2) | 65.29 (5) | | Poker Hand | 82.34 (1) | 75.38 (3) | 67.38 (4) | 64.43 (5) | 78.44 (2) | | Sensor Data | 78.64 (2) | 68.54 (5) | 73.48 (3) | 80.47 (1) | 71.46 (4) | | Average ranks | 1.9 | 3.0 | 3.9 | 2.5 | 3.7 |
|
|
