The week's pick
Random Articles
Reseach Article
Adaptive Cyclic Reconstruction Regularized Bidirectional Feed Forward Network for Robust Malware Classification
| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 187 - Number 87 |
| Year of Publication: 2026 |
| Authors: Nagaraj Shet Manjappa Narahari, Jagadeesha Ramegowda |
10.5120/ijca2026926519
|
Nagaraj Shet Manjappa Narahari, Jagadeesha Ramegowda . Adaptive Cyclic Reconstruction Regularized Bidirectional Feed Forward Network for Robust Malware Classification. International Journal of Computer Applications. 187, 87 ( Mar 2026), 50-56. DOI=10.5120/ijca2026926519
Abstract
The high structural similarity of the variants and the existence of the subtle visual pattern in binary representations have made malware classification more difficult than ever, as the malware families evolve very quickly. Despite the promising performance of deep learning-based image malware classification systems, the majority of the currently used methods are based on fixed loss functions, which do not respond to changing learning dynamics in the course of training, resulting in unsteady feature representations and poor generalization. In order to overcome these issues, this paper presents an Adaptive Cyclic Reconstruction-Regularized Bidirectional Feed Forward neural network, A-CRCL-BFFNN, to classify malware families strongly. The framework proposed is built upon the concept of a bidirectional contrastive autoencoder, in combination with a dynamic loss regulation scheme that balances dynamically between the reconstruction loss, cyclic consistency loss and contractive regularization through training behavior. Such an adaptation provides the model with the ability to learn structural preservation during initial training stages and increasingly implement robustness and latent space smoothness with each succeeding training stage. Binaries of malware are then converted to color-mapped image representation and normalized to maximize the discriminative features learning. In-depth experiments over the Malimg and BIG2015 benchmark datasets revealed that the presented A-CRCL-BFFNN has the highest classification accuracy, recall, and F1-score, with the incidental computation overhead.
References
- J. Ferdous, R. Islam, A. Mahboubi, and M. Z. Islam, “A survey on ML techniques for Multi-Platform Malware Detection: securing PC, mobile devices, IoT, and cloud environments,” Sensors, vol. 25, no. 4, p. 1153, Feb. 2025, doi: 10.3390/s25041153.
- A. Hussain, A. Saadia, M. Alhussein, A. Gul, and K. Aurangzeb, “Enhancing ransomware defense: deep learning-based detection and family-wise classification of evolving threats,” PeerJ Computer Science, vol. 10, p. e2546, Nov. 2024, doi: 10.7717/peerj-cs.2546.
- M. Ashawa, N. Owoh, S. Hosseinzadeh, and J. Osamor, “Enhanced Image-Based malware classification using Transformer-Based Convolutional Neural Networks (CNNs),” Electronics, vol. 13, no. 20, p. 4081, Oct. 2024, doi: 10.3390/electronics13204081.
- F. Harrou, B. Bouyeddou, A. Dairi, and Y. Sun, “Exploiting Autoencoder-Based anomaly detection to enhance cybersecurity in power grids,” Future Internet, vol. 16, no. 6, p. 184, May 2024, doi: 10.3390/fi16060184.
- F. S. Alsubaei, A. A. Almazroi, W. S. Atwa, A. A. Almazroi, N. Ayub, and N. Z. Jhanjhi, “Adaptive malware identification via integrated SimCLR and GRU networks,” Scientific Reports, vol. 15, no. 1, p. 25309, Jul. 2025, doi: 10.1038/s41598-025-08556-4.
- B. Thiyam and S. Dey, “An improved deep autoencoder-based network intrusion detection system with enhanced performance,” International Journal of Internet Technology and Secured Transactions, vol. 13, no. 3, pp. 270–290, Jan. 2024, doi: 10.1504/ijitst.2024.136658.
- S. Li, J. Wang, S. Wang, and Y. Song, “PAFE: A lightweight visualization-based fast malware classification method,” Heliyon, vol. 10, no. 16, p. e35965, Aug. 2024, doi: 10.1016/j.heliyon.2024.e35965.
- Q. Shi et al., “TransGraphNet: robust detection of malicious encrypted network traffic via transformer and graph neural models,” PeerJ Computer Science, vol. 11, p. e3353, Nov. 2025, doi: 10.7717/peerj-cs.3353.
- H. KauserSk and M. AnuV, “Hybrid deep learning model for accurate and efficient android malware detection using DBN-GRU,” PLoS ONE, vol. 20, no. 5, p. e0310230, May 2025, doi: 10.1371/journal.pone.0310230.
- F. A. Abdulazeez, I. T. Ahmed, and B. T. Hammad, “Examining the performance of various pretrained convolutional neural network models in malware detection,” Applied Sciences, vol. 14, no. 6, p. 2614, Mar. 2024, doi: 10.3390/app14062614.
- S. Singh, D. Krishnan, V. Vazirani, V. Ravi, and S. A. Alsuhibany, “Deep hybrid approach with sequential feature extraction and classification for robust malware detection,” Egyptian Informatics Journal, vol. 27, p. 100539, Sep. 2024, doi: 10.1016/j.eij.2024.100539.
- W. Yan, J. Tang, and S. Stucki, “Design and implementation of a lightweight deep CNN-Based plant biometric authentication System,” IEEE Access, vol. 11, pp. 79984–79993, Jan. 2023, doi: 10.1109/access.2023.3296801.
- F. Wang et al., “MalSort: Lightweight and efficient image-based malware classification using masked self-supervised framework with Swin Transformer,” Journal of Information Security and Applications, vol. 83, p. 103784, May 2024, doi: 10.1016/j.jisa.2024.103784.
- J. Yang, J. Cui, and M. Huang, “Extreme Weather Prediction for Highways Based on LSTM-CNN,” 5th International Conference on Applied Machine Learning (ICAML), pp. 191–196, Jul. 2023, doi: 10.1109/icaml60083.2023.00045.
- R. Song, J. Yv, and Z. Du, “Test analysis and optimization of valve flow characteristic curve of 660MW unit,” 2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC), vol. 53, pp. 255–258, Dec. 2022, doi: 10.1109/imcec55388.2022.10019839.
- S. Andriani, S. Galantucci, A. Iannacone, A. Maci, and G. Pirlo, “CNN-AutoMIC: Combining convolutional neural network and autoencoder to learn non-linear features for KNN-based malware image classification,” Computers & Security, vol. 156, p. 104507, May 2025, doi: 10.1016/j.cose.2025.104507
- C. Chen, Y. Hu, L. Gao, and B. Li, “An Optimization Method for Impedance Matching Teardrop via Pads Based on TDR Simulation,” International Applied Computational Electromagnetics Society Symposium (ACES-China), pp. 1–2, Aug. 2023, doi: 10.23919/aces-china60289.2023.10249876.
Index Terms
Keywords