Fig. 3 | Scientific Reports

Fig. 3

From: A multi-label visualisation approach for malware behaviour analysis

Fig. 3The alternative text for this image may have been generated using AI.

Bayesian Grad-CAM and Grad-CAM visualisations exhibit notable differences. By applying LIME to identify and highlight the regions most influential in the model’s prediction for a given class label, it is observed that Bayesian Grad-CAM primarily aligns with these key regions, indicating a closer correspondence to the features driving the model’s decision.

Back to article page