摘要
机器学习算法为很多安全应用提供了良好的解决方案,然而机器学习算法本身却面临被敌手攻击的威胁。为分析敌手攻击对机器学习算法造成的影响,本文提出符合某些特定场合的敌手攻击模型,并在该模型下比较几种线性分类器的对抗性。最后在垃圾邮件过滤公开数据库上进行测试,实验结果表明,支持向量分类器具有相对较好的对抗性。
Machine learning algorithms provide a well solution for many security applications. Machine learning algorithms them- selves, however, face the thread of adversary attack. In order to analyze the impact of adversary attacks on machine learning algo- rithms, the paper presents an adversary attack model in line with some actual situations and compares the antagonism of some lin- ear classifier under this model. The performances of these adversarial classifiers are evaluated on a large public spam corpus. The experiment results show that SVM is more antagonism than other linear classifiers.
出处
《计算机与现代化》
2012年第3期78-81,共4页
Computer and Modernization
关键词
对抗分类
线性分类器
垃圾邮件过滤
对抗性
adversary classification
linear classifier
spam filter
antagonism