欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对当前入侵检测中存在检测率低,误检率和漏报率高的问题,提出了一种基于K-means聚类的贝叶斯分类算法（IKMNB）。用改进的K-means聚类算法对原始数据集中的完整数据进行聚类,然后计算缺失数据集中的每条记录与k个簇中心之间的近似度距离,把记录归属为距离最近的一个簇,使得该记录的缺失值被相应簇中的属性值填充,最后运用贝叶斯分类算法对处理后的完整数据集进行分类。通过仿真实验验证了该算法与朴素贝叶斯算法相比提高了检测率,降低了误检率和漏报率。
英文摘要：
Aiming at the low detection rate and the high rate of error rate as well as omission rate in current intrusion detection,this paper proposed a Bayesian classification algorithm based on K-means clustering. Firstly,the paper improved the clustering algorithm by using the improved K-means algorithm to cluster the complete data in the set of the original data. Secondly,the algorithm calculated the approximate distance between each record of the missing dataset and k cluster centers,and took the record belongs to the nearest cluster,and the missing value of the record was filled with property value of the corresponding cluster. Finally,it used the Bayesian classification algorithm to classify the processed complete data set. The simulation experiments prove that compared with the nave Bayesian,the detection rate is improved and the false detection rate and the omission rate are lower by the improved algorithm.
同期刊论文项目
同项目期刊论文
期刊信息
位置：