Deep Learning Based Dynamic Malware Detection Framework

Malware is a significant threat to the security and stability of computer systems and networks.Traditional signature-based methods for detecting malware are no longer sufficient due to cyberattacks’ increasing number and sophistication.Therefore,there is a need for more advanced and effective techniques for detecting malware in real-time.Deep learning has emerged as a powerful tool for detecting malware,but it requires high-quality features to achieve accurate results.Feature engineering plays a critical role in the effectiveness of deep learning models.In this study,we propose a dynamic malware detection framework based on deep learning and feature engineering.The proposed consists of three stages: preprocessing,feature extraction and classification.In the preprocessing stage,we normalize and resize the images of malware samples to a fixed size.In the feature extraction stage,we use a convolutional neural network(CNN)to extract high-level features from the preprocessed image.We also propose a feature selection method based on mutual information to select the most relevant features for classification.In the classification phase,we use support vector machines(SVM)to classify malware samples into their respective families.We evaluate the performance of the proposed framework using the NSLKDD dataset,which consists of 9,997 malware samples belonging to 25 families.We compare the proposed framework with existing techniques such as PCA-SVM,CNN,and deep autoencoders.We evaluate the performance of the proposed framework using various metrics such as accuracy,precision,recall and F1-score.Experimental results show that the deep learning-based framework using feature engineering to detect cyber-attacks achieves superior performance compared to traditional machine learning algorithms and other deep learning-based models.The framework’s high accuracy,precision,recall,and F1 scores demonstrate its effectiveness in detecting cyberattacks in network traffic data,which is critical for ensuring Cyber security.The proposed framework achieves high accuracy,precision,and recall,outperforming the state-of-the-art in detecting malware.The proposed framework achieves 97.91% precision,97.95% accuracy,and 97.90%recall.Mutual information-based feature selection methods also improve the performance of the proposed framework.