Malware Classification Using Cyclic Reconstruction Contractive Loss based Bidirectional Feed Forward Neural Network.
Saved in:
| Title: | Malware Classification Using Cyclic Reconstruction Contractive Loss based Bidirectional Feed Forward Neural Network. |
|---|---|
| Authors: | Manjappa Narahari, Nagaraj Shet, Ramegowda, Jagadeesha |
| Source: | International Journal of Intelligent Engineering & Systems; 2026, Vol. 19 Issue 1, p114-127, 14p |
| Subject Terms: | MALWARE, FEEDFORWARD neural networks, FEATURE extraction, PATTERN perception, LOSS functions (Statistics), STATISTICAL reliability, ARTIFICIAL neural networks, MACHINE learning |
| Abstract: | Malware classification focuses on grouping malicious software depending on similarities in patterns and structure. It enables distinguishing distant types of malwares effectively by utilizing learned features from representations. However, classifying malware remains challenging due to high similarity among variants and presence of complex structural patterns, which can lead to inaccurate classification. In this research, Cyclic Reconstruction Contractive Loss based Bidirectional Contrastive Autoencoder Feed Forward Neural Network (CRCL-BFFNN) is proposed to classify malware accurately. In traditional FFNN, bidirectional contrastive autoencoder is incorporated, which allows capturing both backwards and forward relationships in malware data and enhances feature richness. Bidirectional contrastive autoencoder preserves significant structural information while ensuring consistency over encoding and decoding, which leads to more accurate and robust malware classification. CRC loss function assists the network for learning stable and consistent representation which enforces local smoothness as well as robustness across different malware samples. Hence, proposed CRCL-BFFNN achieves a high accuracy of 99.71% and 99.78% on BIG2015 and Malimg dataset compared to existing methods like MalSort. [ABSTRACT FROM AUTHOR] |
| Copyright of International Journal of Intelligent Engineering & Systems is the property of Intelligent Networks & Systems Society and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Complementary Index |
Nájsť tento článok vo Web of Science Be the first to leave a comment!