Table 13. Comparison of GN-BiLSTM with other state-of-the art work on same dataset.

Reference	Technique	Validation	Malware	Category	Family
Roy et al. (2023)	ML: Stacked ensemble	4-K cross-validation	99.98%	85.04%	70.29%
Abualhaj et al. (2024)	ML: KNN	–	99.97%	82.21%	66.93%
Mezina & Burget (2022)	DL: Dilated CNN	–	99%	–	83%
Roy & Chen (2021)	DL: TF-IDF	–	99.87%	–	96.5%
Shafin et al. (2023)	–	–	99.96%	84.56%	72.60%
Proposed GN-BiLSTM	DL: Feature engineering & SMOTE	K-Fold Cross-validation	99.99%	85.48%	74.68%