Journal of Electrical and Computer Engineering / 2020 / Article / Tab 6 / Research Article
A Classification and Novel Class Detection Algorithm for Concept Drift Data Stream Based on the Cohesiveness and Separation Index of Mahalanobis Distance Table 6 Experimental result data in ArtificialCDS dataset.
C C&NCBM accuracy (%) MineClass accuracy (%) KNN accuracy (%) C&NCBM evaluation time (s) MineClass evaluation time (s) KNN evaluation time (s) 1 76.0900 74.8900 74.1500 7.7969 7.3750 7.3088 2 76.2050 75.2600 74.4200 16.7031 15.0781 14.6625 3 76.3500 75.7800 74.6300 25.2500 22.8438 21.9838 4 76.3775 75.8050 74.5575 34.2031 30.4844 29.3313 5 76.4500 75.8640 74.4540 44.4844 38.2344 36.6325 6 76.5467 75.9917 74.4083 53.9219 45.9063 44.2265 7 76.5686 75.9557 74.4586 63.6094 53.6250 51.7875 8 76.6013 75.9388 74.4088 73.2969 61.2969 59.4813 9 76.6444 76.0044 74.3711 82.1719 69.0625 67.0588 10 76.6870 75.9800 74.3850 90.7500 76.7188 74.5263