Advanced Deep Learning Approaches for Malware Detection in Cybersecurity Datasets

Asadi Srinivasulu* and Gaythri
Professor of CSE and Head of R & D, Sree Dattha Group of Educational Institutions, Hyderabad, India
*Corresponding Author: Asadi Srinivasulu, Professor of CSE and Head of R & D, Sree Dattha Group of Educational Institutions, Hyderabad, India.

Published: October 10, 2023

View Pdf

Abstract  

As cyber-attacks multiply rapidly and malware becomes more intricate, conventional signature-focused detection methods find it challenging to keep up. In light of this, deep learning approaches have emerged as a potent solution to boost malware detection efficacy. This study delves into sophisticated deep-learning strategies to spot and categorize malware in modern cybersecurity data. We present a combined model that merges the strengths of Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN), harnessing both the spatial and sequential attributes of the data. When pitted against classic deep learning setups, our model showcases remarkable enhancements in detection precision, marking a 98.7% success rate and a mere 0.5% rate of false positives using the XYZ dataset. Additionally, we deploy specific data augmentation methods designed for cybersecurity data, augmenting our model's adaptability and resilience to new threats. The results indicate that employing cutting-edge deep learning designs can fortify malware detection mechanisms, providing superior defense in a constantly shifting cyber-threat environment.

Keywords: Deep Learning; Malware Detection; Signature-based Detection; Convolutional Neural Networks(CNN); Recurrent Neural Networks (RNN) and Data Augmentation

.