文章提出了基于粗糙集属性约简的支持向量异常入侵检测方法。为验证该方法的有效性,对实验数据集KDD99分别用粗糙集属性约简的支持向量分类方法和传统的支持向量分类方法进行实验仿真,并把两者的实验结果进行对比。实验证明,基于粗糙集属性约简的支持向量异常入侵检测方法在检测精度相当的情况下,有效的降低了检测时间并减少了存储空间。 This paper presents a support vector anomaly detection method based on rough set attribute reduction. In order to verify the effectiveness of the proposed method, the experiment data set KDD99 is respectively simulated by the support vector classification method of the rough set attribute reduction and the traditional support vector classification method, and the experimental results of the two are compared. Experiments show that the support vector anomaly detection method based on rough set attribute reduction can effectively reduce the detection time and reduce the storage space under the same detection accuracy.