Title: Leveraging Secondary Feature Fusion for a Hybrid Malware Classification Strategy

The exponential growth in the volume of malware, encompassing both new variants and previously unseen strains, has rendered detection and mitigation increasingly difficult. While automating the detection and categorization of malware into specific families can enhance mitigation efforts, conventional detection techniques often fail to assign identified threats to their correct families, thereby limiting their utility. To address this gap, this study introduces a framework that simultaneously automates malware detection and classifies the identified samples into their respective families.

The methodology relies on a customized feature selection process applied to extracted attributes, including API calls and both fixed and variable length n-grams. These features are combined through a fusion technique. For the prediction phase, the system employs an algorithm fusion strategy based on voting mechanisms.

To validate the approach, the researchers utilized a dataset provided by Microsoft, applying both binary and multi-class classification models. The results were benchmarked against current state-of-the-art methods. The findings demonstrate that the proposed hybrid approach is both effective and efficient, achieving an Area Under the Curve (AUC) of 0.989, an accuracy rate of 99.72%, and a log loss of 0.01.

Source: arXiv Generated at: 2026-06-03 00:00:00 UTC