Table 3 Binary Classification Performance Analysis Between DECODE and Baseline Methods. The metrics include ACC (Accuracy), FPR (False Positive Rate), FNR (False Negative Rate), and MCC (Matthews Correlation Coefficient). All values range from 0 to 1.
From: A multi-label visualisation approach for malware behaviour analysis
Model | ACC | F1 Score | FNR | FPR | MCC |
|---|---|---|---|---|---|
DECODE | 0.9380 | 0.9619 | 0.0655 | 0.0432 | 0.8032 |
InceptionResNetV2 | 0.9200 | 0.9599 | 0.0343 | 0.2486 | 0.7370 |
ResNet18 | 0.9170 | 0.9562 | 0.0301 | 0.1910 | 0.7023 |
InceptionV4 | 0.9170 | 0.9441 | 0.0484 | 0.1892 | 0.7414 |
ResNet34 | 0.9140 | 0.9568 | 0.0479 | 0.2171 | 0.7181 |
VGG16 | 0.9060 | 0.9520 | 0.0585 | 0.1946 | 0.7137 |
ResNet50 | 0.8880 | 0.9455 | 0.0872 | 0.1711 | 0.5951 |
