Scholarlink

This study proposes a method for categorizing populations of malicious softwareMalicious software using an operation code adjacency graphOperation code adjacency graph feature. To extract an operation code sequence and build an operation code adjacency graphOperation code adjacency graph from it, a piece of input software must be deconstructed. Then, a feature vectorFeature vector is built using this graph. To build a model for hazardous software group classification, a deep neural networkDeep neural network is trained utilizing feature vectorsFeature vector from well-known group classifications. The family group of unidentified software samples is categorized using the learned model. The poor accuracy that plagues present techniques for identifying dangerous software is addressed by the suggested strategy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Neural Network-Based Classification of Spyware Populations Using Macroscopic Statistical Attributes of Operation Codes

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Sonal Malhotra,
  • Deepak Upadhyay,
  • Rahul Chauhan,
  • Arun Balodi,
  • Swati Devliyal

摘要

This study proposes a method for categorizing populations of malicious softwareMalicious software using an operation code adjacency graphOperation code adjacency graph feature. To extract an operation code sequence and build an operation code adjacency graphOperation code adjacency graph from it, a piece of input software must be deconstructed. Then, a feature vectorFeature vector is built using this graph. To build a model for hazardous software group classification, a deep neural networkDeep neural network is trained utilizing feature vectorsFeature vector from well-known group classifications. The family group of unidentified software samples is categorized using the learned model. The poor accuracy that plagues present techniques for identifying dangerous software is addressed by the suggested strategy.