Fig 1.
Classification of Malware [1].
Table 1.
Performance comparison of state-of-art malware classification existing techniques.
Fig 2.
Flow Diagram of Proposed model.
Table 2.
Grouping the malware classes and sub-classes.
Fig 3.
The number of samples for each malware class.
Table 3.
Encoding API and behavior features types, and description.
Fig 4.
Feature Importance Scores for Malware Behavior Classification.
Table 4.
Final performance metrics after 100 epochs.
Table 5.
Malware families and class-specific performance measure.
Fig 5.
Correlation matrix of performance metrics calculated from class-specific.
Table 6.
Top five highest and lowest performing classes based on their F1-Scores.
Table 7.
Group-wise malware classification average results.
Table 8.
Per-Classifier Performance by Class Category.
Fig 6.
Graphical representation of per-classifier by class category.
Table 9.
BERT features vector classification results on different classifiers.
Table 10.
Comparative Analysis of BERT features with Manual features on different classifiers.
Table 11.
Comparison of the proposed model with other state-of-the-art Techniques.