Table 5 Comparison of Different Algorithms’ Performance on Classification Tasks in DGA Data.
Method type | Algorithm | Data type | Accuracy (%) | Remarks |
|---|---|---|---|---|
Mechanistic model | Non-parametric kernel density | Raw Data of DAG | 64 | / |
Gaussian density | Raw Data of DAG | 54.4 | / | |
Fuzzy set | Raw Data of DAG | 53.8 | / | |
Machine learning | Random Forest | Raw Data of DAG | 88 | / |
AdaBoost | Raw Data of DAG | 56 | / | |
KNN | Raw Data of DAG | 65 | / | |
BP neural network | Raw Data of DAG | 40 | / | |
Logistic regression | Raw Data of DAG | 67 | / | |
Deep network | CNN | DGA Feature Image | 81 | Epoch = 5 |
CNN | DGA Feature Image | 95 | Epoch = 10 | |
EfficientNet-B0 | DGA Feature Image | 98 | Epoch = 5 | |
ResNet-18 | DGA Feature Image | 98 | Epoch = 5 | |
DETR + X | DGA Feature Image | 100 | Epoch = 5 (The classifier is Vit) |
