Enhancing Cyber Security Defences through Accurate Malware Classification

The rapid integration of virtual environments into human lives, accelerated by the COVID-19 pandemic, has shifted criminal activities to the digital realm, where malware serves as a primary tool for cybercriminals. Conventional detection methods struggle against the evolving sophistication of malware. To address this challenge, a novel approach leveraging deep learning through the Customized Deep Learning-based Malware Classification (CDL-MC) architecture is proposed for detecting new and complex malware types. Objectives include developing an image-based malware dataset and implementing CDL-MC with multiple convolutional and fully connected layers. Additionally, the model is deployed in a user-friendly MATLAB GUI application to enhance practical usability in cybersecurity. Experimental results demonstrate the efficacy of the method in classifying malware with high accuracy, surpassing existing state-of-the-art methods. By harnessing deep learning, the approach offers a promising solution to combat ever-evolving malware threats and ensure enhanced cybersecurity in the virtual age.