摘要
提出一种支持向量机的木马检测方法。首先对木马程序进行特征提取,然后采用支持向量机对木马数据进行训练,建立一个木马检测模型,最后采用木马检测模型对检测程序进行检测,将其分类为合法程序或木马程序。实验表明,采用本方法木马检测准确率在98,4%,远远高于传统木马检测算法的准确正确率,且检测时间更少,更加符合木马实时性要求高的特点。
This paper ptlt forward a Trojan detection method based on support vector machine. The Trojan program features are extracted, and then use support vector machine to train Trojan data to build a Trojan detection model, finally using the Trojan horse detection model treats the detection procedure for detection, classified as a legitimate or Trojan program. This method of Trojan detection accuracy rate is high at 98, 4%, much higher than the traditional Trojan detection algorithm accurately correct rate
出处
《科技通报》
北大核心
2012年第2期39-41,共3页
Bulletin of Science and Technology
基金
湖南省自然科学基金项目(11JJ6056)
关键词
木马病毒
模式识别
支持向量机
trojan horse
pattern recognition
support vector machines
