Stacking Ensemble-Based Approach for Malware Detection

Stacking Ensemble-Based Approach for Malware Detection

연구 분야: Safety

학회: SN Computer Science

The rapid growth of Internet connectivity has resulted in a significant increase in digital attack events, many of which have devastating and severe consequences. Malware is one type of cyber attack that is becoming more common by the day. With the rapid evolution of malware as technological innovation advances, the battle between security researchers and malware developers is ongoing. Analysts are working to distinguish it, while cyber criminals are figuring out how to hide it. Many researchers have proposed various methods for detecting malware, of which memory analysis plays a vital role. In this study, an efficient stack-based detection approach is proposed by combining kNN, Random Forest, Neural Network, Gradient Boosting, and Adaboost learning algorithms to detect the malware more efficiently. The proposed model is the more complex by combining the five approaches into two learning layers to classify the instanced more accurately. As per the result obtained, the proposed approach achieved high accuracy during training and testing phases using the memory forensic malware dataset.