Research Article
CFSBFDroid: Android Malware Detection Using CFS + Best First Search-Based Feature Selection
Table 9
Performance of various classifiers.
| Classifier | Class | TP rate | FP rate | Precision | Recall | MCC | ROC area | F-measure |
| RF | 0 | 0.987 | 0.032 | 0.969 | 0.987 | 0.956 | 0.995 | 0.978 | 1 | 0.968 | 0.013 | 0.987 | 0.968 | 0.956 | 0.995 | 0.977 | Weighted Avg. | 0.978 | 0.022 | 0.978 | 0.978 | 0.956 | 0.995 | 0.978 |
| REPTree | 0 | 0.993 | 0.063 | 0.941 | 0.993 | 0.931 | 0.981 | 0.966 | 1 | 0.937 | 0.007 | 0.992 | 0.937 | 0.931 | 0.981 | 0.964 | Weighted Avg. | 0.965 | 0.035 | 0.966 | 0.965 | 0.931 | 0.981 | 0.965 |
| Rule PART | 0 | 0.987 | 0.053 | 0.949 | 0.987 | 0.934 | 0.980 | 0.967 | 1 | 0.947 | 0.013 | 0.986 | 0.947 | 0.980 | 0.983 | 0.966 | Weighted Avg. | 0.967 | 0.033 | 0.967 | 0.967 | 0.934 | 0.980 | 0.967 |
| LMT | 0 | 0.999 | 0.000 | 1.000 | 0.999 | 0.999 | 1.000 | 0.999 | 1 | 1.000 | 0.001 | 0.999 | 1.000 | 0.999 | 1.000 | 0.999 | Weighted Avg. | 0.999 | 0.001 | 0.999 | 0.999 | 0.999 | 1.000 | 0.999 |
| SMO | 0 | 0.996 | 0.039 | 0.962 | 0.996 | 0.957 | 0.978 | 0.979 | 1 | 0.961 | 0.004 | 0.996 | 0.961 | 0.957 | 0.978 | 0.978 | Weighted Avg. | 0.978 | 0.022 | 0.979 | 0.978 | 0.957 | 0.978 | 0.978 |
| MCC | 0 | 0.999 | 0.001 | 0.999 | 0.999 | 0.997 | 0.999 | 0.999 | 1 | 0.999 | 0.001 | 0.999 | 0.999 | 0.997 | 0.999 | 0.999 | Weighted Avg. | 0.999 | 0.001 | 0.999 | 0.999 | 0.997 | 0.999 | 0.999 |
| SGD | 0 | 0.997 | 0.018 | 0.982 | 0.997 | 0.979 | 0.989 | 0.989 | 1 | 0.982 | 0.003 | 0.997 | 0.982 | 0.979 | 0.989 | 0.989 | Weighted Avg. | 0.989 | 0.011 | 0.989 | 0.989 | 0.979 | 0.989 | 0.989 |
|
|