中文摘要：

为解决入侵检测领域计算复杂度、时间复杂度高的难题,达到更优秀的入侵检测效果,有效降维,在原有的ReFCBF算法的基础上,提出增强区分特征间互信息的能力,以在改进的Re-ReliefF算法的基础上,实现更佳的入侵检测效果为目标。实验采用DARPA 2000数据集,对数据的41维特征进行选择,采用支持向量机作为分类器,实验结果表明,该改进方法在分类的耗时和误报率略好的情况下,提高了30%的准确率。

英文摘要：

To solve the problem of high computational complexity and time complexity,to achieve more effective intrusion detection results,an effective dimensionality reduction,on the basis of the original Re-FCBF algorithm,was proposed with the ability of distinguishing characteristics between mutual information,and to achieve more effective dimensionality reduction.On the basis of improvements to Re-ReliefF algorithm,achieving better intrusion detection effects was set as the goal.DARPA 2000 data sets were used to select 41-dimensional feature data for the experiment,and support vector machine was used as the classifier.Experimental results show the improved method proposed increases the classification accuracy by 30% on the premise of slightly reducing time-consumption and the rate of false.