基于Win32 API和SVM的未知病毒检测模型
首发时间:2014-11-10
摘要:分析现有的未知病毒检测方法,提出了一种以关键API调用及其动态参数作为特征,并基于SVM支持向量机进行未知病毒检测的新方法。该方法使用关键API并结合其重要参数预先定义恶意事件行为及其影响因子,从而形成多维特征向量,并通过信息熵理论对特征向量进行筛选,将经过降维处理过的特征向量利用支持向量机分类方法进行训练,最终实现对未知病毒的检测。该方法有效的克服了病毒特征码扫描技术无法识别未知病毒的缺点和静态API序列检测法对隐藏API调用的未知病毒的低识别率。实验结果表明该病毒检测方法具有较高的准确率,有一定的应用价值。
For information in English, please click here
Unknown virus detection model based on Win32 API and SVM
Abstract:This paper analyzes the existing unknown virus detection methods, proposing a new method using API calls with dynamic paremeters and SVM-based support vector machine for unknown virus detection. The method uses predefined key APIs in combination with important parameters and their influencing factors to get the multi-dimensional feature vector. The vector which is formed by dimensionality reduction filter uses support vector machine classification training method to detect unknown viruses. The method effectively overcomes the shortage of virus feature code scanning method that cannot identify unknown viruses and the static API sequence detection method that has a low recognition rate with unknown viruses hidden API calls.Experimental results show that the detection method has better classfication accuracy, so there is a certain practical value.
Keywords: information Security support vector machine unknown virus detection
基金:
论文图表:
引用
No.4616294101434014****
同行评议
共计0人参与
勘误表
基于Win32 API和SVM的未知病毒检测模型
评论
全部评论