MALWARE DETECTION VIA CLASSIFICATION

This paper has presented an original workflow for malware detection via classification using CIC-MalMem-2022, six supervised learning algorithms and a Python Flask interface. The methodology emphasizes controlled preprocessing, leakage prevention, consistent evaluation and practical deployment. No experimental winner or accuracy value is asserted because the project's measured outputs were not supplied. Once the implementation is executed, the results table and discussion should be completed using the held-out test metrics. Future work can extend binary detection to classification of ransomware, spyware and Trojan categories; evaluate group-based or time-aware splits; investigate feature importance and explainability; test robustness against distribution shift; and measure prediction latency under realistic request loads. Deployment should be further assessed for input validation, logging, access control and secure model updates.

[1]  M. Dener, G. Ok, and A. Orman, “Malware detection using memory analysis data in big data environment,” Applied Sciences, vol. 12, no. 17, Art. no. 8604, 2022, doi: 10.3390/app12178604. [2]  D. Cevallos-Salas, F. Grijalva, J. Estrada-Jiménez, D. S. Benítez, and R. O. Andrade, “Obfuscated privacy malware classifiers based on memory dumping analysis,” IEEE Access, vol. 12, pp. 17481–17498, 2024, doi: 10.1109/ACCESS.2024.3358840. [3] Scikit-learn Developers, “Metrics and scoring: quantifying the quality of predictions,” scikit-learn documentation. [Online]. Available: https://scikit-learn.org/stable/modules/model_evaluation.html. Accessed: Oct. 30, 2026. [4]  Canadian Institute for Cybersecurity, “CIC-MalMem-2022 dataset,” University of New Brunswick. [Online]. Available: https://www.unb.ca/cic/datasets/malmem-2022.html. Accessed: Oct. 30, 2026.