The study presents a malware detection system with an impressive accuracy of 99.99% in binary classification. Various methodologies were explored, including analyzing original, undersampled, and oversampled datasets. XGBoost classifiers consistently performed best across various metrics, highlighting the challenges posed by dataset imbalance. The research provides valuable insights into effective techniques for malware detection and classification, establishing a foundation for future enhancements in cybersecurity methodologies.
Our malware detection system achieved a remarkable 99.99% accuracy on the test set, successfully identifying all malware instances and demonstrating the effectiveness of our approach.
Despite the challenges posed by an imbalanced dataset, our step-wise experiments have shown significant promise in enhancing both the detection and classification of malware.
Collection
[
|
...
]