Malware Detection Using LC-KSVD with Sparse Random Projection

• Item type: SIG Technical Reports(1)
• 公開日: 2024-08-28
• タイトル: Malware Detection Using LC-KSVD with Sparse Random Projection
• タイトル / 言語: en
• タイトル: Malware Detection Using LC-KSVD with Sparse Random Projection
• 言語: eng
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Graduate School of Engineering and Science, University of the Ryukyus
• 著者所属: Information Technology Center, University of the Ryukyus
• 著者所属(英): en
• 著者所属(英): Graduate School of Engineering and Science, University of the Ryukyus
• 著者所属(英): en
• 著者所属(英): Information Technology Center, University of the Ryukyus
• 著者名: Aziz, Urahman Shafaq; Takayuki, Nakachi
• 著者名(英): Aziz, Urahman Shafaq; Takayuki, Nakachi
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Malware detection is increasingly challenging due to the sophistication of threats. This study presents a small data driven malware detection approach that integrates Label Consistent K-SVD (LC-KSVD) with Sparse Random Projection (SRP). SRP effectively reduces the dimensionality of high-dimensional malware image data while maintaining its core structure, and LC-KSVD, which incorporates label information, refines dictionary learning for better classification. The unified objective function, optimized using K-SVD, combines reconstruction and classification errors. The approach is evaluated based on classification accuracy, computational complexity. Simulation results demonstrate that LC-KSVD with SRP achieves better classification accuracy compared to LC-KSVD with traditional dimensionality reduction methods, potentially advancing the field of malware detection and cybersecurity.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Malware detection is increasingly challenging due to the sophistication of threats. This study presents a small data driven malware detection approach that integrates Label Consistent K-SVD (LC-KSVD) with Sparse Random Projection (SRP). SRP effectively reduces the dimensionality of high-dimensional malware image data while maintaining its core structure, and LC-KSVD, which incorporates label information, refines dictionary learning for better classification. The unified objective function, optimized using K-SVD, combines reconstruction and classification errors. The approach is evaluated based on classification accuracy, computational complexity. Simulation results demonstrate that LC-KSVD with SRP achieves better classification accuracy compared to LC-KSVD with traditional dimensionality reduction methods, potentially advancing the field of malware detection and cybersecurity.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AN10438399
• 書誌情報: 研究報告オーディオビジュアル複合情報処理（AVM）; 巻 2024-AVM-125, 号 25, p. 1-6, 発行日 2024-08-28
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8582
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会