Research Article
HYBRID-CNN: An Efficient Scheme for Abnormal Flow Detection in the SDN-Based Smart Grid
Table 4
Performance comparison of the proposed and state-of-the-art methods.
| Reference | Method | Proportion = 80% | Proportion = 70% | Proportion = 60% | Acc | DR | FPR | Acc | DR | FPR | Acc | DR | FPR |
| Ashraf et al. [17] | Naive Bayes | 0.7663 | 0.8514 | 0.3841 | 0.7669 | 0.8611 | 0.3999 | 0.7655 | 0.8512 | 0.3883 | Reddy et al. [19] | SVM | 0.7594 | 0.6895 | 0.1170 | 0.7257 | 0.7806 | 0.3714 | 0.7346 | 0.7874 | 0.3591 | Xin et al. [22] | LSTM | 0.8916 | 0.9843 | 0.2724 | 0.8897 | 0.9840 | 0.2775 | 0.8894 | 0.9835 | 0.2778 | Wang et al. [25] | CNN-LSTM | 0.8995 | 0.9612 | 0.2095 | 0.8965 | 0.9460 | 0.1910 | 0.8955 | 0.9571 | 0.2138 | Proposed method | HYBRID-CNN | 0.9564 | 0.9856 | 0.0442 | 0.9408 | 0.9382 | 0.0544 | 0.9386 | 0.9493 | 0.0803 |
|
|