欢迎访问《空军工程大学学报》官方网站!

咨询热线:029-84786242 RSS EMAIL-ALERT
CasKDNet:基于改进DenseNet的恶意代码分类方法
DOI:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TP309

基金项目:

国家自然科学基金(61806219,61703426,61876189);陕西省高校科协青年人才托举计划(20190108,20220106);陕西省创新能力支撑计划(2020KJXX-065)


CasKDNet: A Malware Classification Method Based on Improved DenseNet
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对现有恶意代码可视化分类模型在精度和鲁棒性方面的不足,提出一种基于改进DenseNet的恶意代码可视化分类方法CasKDNet,通过3项关键技术实现精度和鲁棒性的提升。首先,构建级联分类器结构,增强纹理相似家族的特征区分能力;其次,采用KAN结构替代DenseNet网络中的多层感知机,优化特征提取过程的非线性表达能力,提升模型整体精度;最后,基于FFM图像修复算法对训练集进行数据增强, 提高模型鲁棒性。在恶意代码数据集Malimg上的实验结果显示,CasKDNet模型取得99.69%的分类准确 率,与现有研究方法相比具有明显性能优势。此外,在白盒攻击背景下,FGSM和I-FGSM算法对CasKDNet的攻击成功率仅为12.7%和37.5%,进一步证实了模型在防范对抗性攻击方面的有效性。

    Abstract:

    In existing malware visualization classification models, there are inadequate accuracy and robustness. For this reason, this paper proposes a malicious code visualization classification method CasKD Net (Cascade DenseNet with KAN) based on an improved DenseNet. The CasKDNet is to realize the improvements in accuracy and robustness by three key technologies. Firstly, a cascaded classifier structure is constructed to enhance the feature discrimination ability of texture similar families. Secondly, the KAN structure is used to replace the multi-layer perceptron in the DenseNet network, optimizing the non-linear expression ability of the feature extraction process and improving the overall accuracy of the model. Finally, the FFM image restoration algorithm is used to enhance the training set and improve the robustness of the model. It appears from the experimental results on the malicious code dataset Malimg that the CasKDNet model achieves 99.69% of classification accuracy, and is superior to the existing research methods. Furthermore, in the context of white box attacks, the success rate of FGSM and I-FGSM algorithms attack against the CasKDNet only serves as 12.7% and 37.5% respectively, and the model is valid in preventing adversarial attacks.

    参考文献
    相似文献
    引证文献
引用本文

刘 强, 王 坚, 路艳丽, 王艺菲. CasKDNet:基于改进DenseNet的恶意代码分类方法[J].空军工程大学学报,2025,26(4):110-119

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2025-08-07
  • 出版日期: