Identification of malware using CNN and bio-inspired technique
N. Poonguzhali, T. Rajakamalam, S. Uma, Manju R A · 2019 IEEE International Conference on System, Computation, Automation and Networking (ICSCAN) · 2019
The growth of the Internet, has dramatically increased the malicious code attacks, with malicious code variants ranking as a key threat to Internet security. The ability to detect variants of malicious code is critical for protection against security breaches, data theft, and other dangers. Both the existing and the current methods of identifying the malicious codes give a low accuracy and takes more time to process. This paper proposes a method that uses deep learning with Convolutional Neural Network for the detection of malware variants. In the current generation deep learning is playing a vital role for predictive analysis. Here we have converted the malicious codes into a gray scale images. The identification and the feature extraction is done by the Convolutional Neural Network and the affected malware images are classified using the Support Vector Machine classifier. The classifier also mentions the malware family that the affected code belongs to. In addition to this we utilized a bio-inspired optimization technique to deal with the imbalance of the data. The experimental results that our model achieved give good accuracy and speed as compared to other malware detection models.