Create robust malware detection system using machine learning and deep learning approaches
In this project, we aimed to investigate the performance of classical machine learning and deep learning methods for malware detection. To this end, we used two provided datasets: one for training and validation and one for testing. The models were trained to classify software samples as either benign or malicious. The results of the analysis revealed that both the machine learning and deep learning models performed quite well in detecting malware. Furthermore, we evaluated the models’ robustness by generating adversarial samples using genetic algorithms such as GAMMA and found that some models were able to generalize, although with slightly lower performance, even in the presence of these samples. The results of this study demonstrate the potential of using machine learning and deep learning techniques for malware detection. Additionally, the use of adversarial samples generated with the genetic algorithms can be an effective method to evaluate the robustness of the models and to verify the transferability of malware obfuscation across heterogeneous models in terms of architecture.