Light-Weight Deep Learning Models for Visual Malware Classification

Abstract

Malware attacks are on the rise every day in the Internet-based digital world. Regular Internet users are at risk due to the evolution of new infections. In recent years, the use of machine learning algorithms to identify malware has gained popularity because numerous studies have demonstrated its efficacy. This work provides two deep learning models to categorize the malware turned into images. Our method uses fewer resources and takes less time to accomplish the same performance as state-of-the-art results. The primary advantage of malware images is that no additional feature engineering is required. Our models for categorizing image-based malware are less complex and can be used in computational systems with limited computational capabilities, such as Android devices. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.