基于增量支持向量机的入侵检测算法

摘要

针对网络入侵检测无法识别新的入侵行为,利用增量学习不断完善分类器,使得分类器可以识别新的入侵行为.提出一种基于相似度的增量支持向量机算法,该算法依据新增样本与支持向量之间的相似度来选择样本(当前分类器缺少该样本的空间信息),然后加入训练集中参加下一次迭代训练.实验结果表明,该算法能够提高最终分类器的分类精度和算法的训练速度.%Because network intrusion detection system cannot identify new intrusion behavior,incremental learning can improve classifier,and classifier can identify the new intrusion behavior.This paper presents an incremental support vector machine algorithm based on similarity.The algorithm selects the sample according to the similarity between the new samples and support vectors.These samples contain the spatial information that the current classifier lacks.Then these samples are added to the training center for the next iteration.This algorithm can improve the training speed and classification accuracy.Finally,the experimental results show that the proposed algorithm can improve the classification accuracy and the training speed.