Mal class: a deep learning approach for automatic classification of malware images
Mal class: a deep learning approach for automatic classification of malware images
These days, malware evolves and multiplies exponentially through structural changes and camouflage using methods like encryption, obfuscation, polymorphism, and metamorphism. As deep learning has advanced, techniques like convolutional neural networks (CNN) have become powerful instruments for identifying complex patterns in this malicious software. The present study leverages CNN's capacity to detect patterns in malware datasets generated from RGB or images in greyscale and to determine the global structure of code that has been converted into an image. Convolutional Neural Networks (CNN) are a method of deep learning that has recently demonstrated better performance than conventional learning algorithms, particularly in applications like image categorization. Motivated by this result, a CNN-based malware sample categorisation architecture is proposed. After converting binaries of malware to monochrome images, we train a CNN to classify the images.
Convolution Neural Network, Deep Learning, Maling dataset, Malware Image Classification, Malware visualisation
1329-1333
Divya, S.
3ee4e63f-4f55-41da-80ae-18de34842645
18 February 2025
Divya, S.
3ee4e63f-4f55-41da-80ae-18de34842645
Divya, S.
(2025)
Mal class: a deep learning approach for automatic classification of malware images.
In 4th International Conference on Sentiment Analysis and Deep Learning, ICSADL 2025 - Proceedings.
IEEE.
.
(doi:10.1109/ICSADL65848.2025.10933410).
Record type:
Conference or Workshop Item
(Paper)
Abstract
These days, malware evolves and multiplies exponentially through structural changes and camouflage using methods like encryption, obfuscation, polymorphism, and metamorphism. As deep learning has advanced, techniques like convolutional neural networks (CNN) have become powerful instruments for identifying complex patterns in this malicious software. The present study leverages CNN's capacity to detect patterns in malware datasets generated from RGB or images in greyscale and to determine the global structure of code that has been converted into an image. Convolutional Neural Networks (CNN) are a method of deep learning that has recently demonstrated better performance than conventional learning algorithms, particularly in applications like image categorization. Motivated by this result, a CNN-based malware sample categorisation architecture is proposed. After converting binaries of malware to monochrome images, we train a CNN to classify the images.
This record has no associated files available for download.
More information
Published date: 18 February 2025
Additional Information:
Publisher Copyright:
© 2025 IEEE.
Venue - Dates:
4th International Conference on Sentiment Analysis and Deep Learning, ICSADL 2025, , Bhimdatta, Nepal, 2025-02-18 - 2025-02-20
Keywords:
Convolution Neural Network, Deep Learning, Maling dataset, Malware Image Classification, Malware visualisation
Identifiers
Local EPrints ID: 503047
URI: http://eprints.soton.ac.uk/id/eprint/503047
PURE UUID: 8a310419-92fd-4a4e-8324-6ab89226539e
Catalogue record
Date deposited: 17 Jul 2025 16:54
Last modified: 18 Jul 2025 02:17
Export record
Altmetrics
Contributors
Author:
S. Divya
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics