底层网络潜在危险数据的未激活状态挖掘模型

Mining Model of Inactive State for Underlying Network Potential Risk Data

底层网络潜在的危险未激活状态数据属性具有较高的模糊性,传统的数据挖掘方法依据攻击行为特征,完成潜在危险数据未激活状态的挖掘,容易产生挖掘多中心问题,具有较高的误警率,提出一种基于动态映射的潜在危险数据未激活状态挖掘方法,面向底层高维度潜在危险数据属性,通过主成分分析方法,对底层网络潜在危险数据进行降维操作,构建潜在危险数据的映射网络,搜索优秀的显示邻域保留的网络,从该网络中生成映射图,对其参数进行分析,获取异常数据,也就是未激活状态数据。实验结果说明,采用所提方法可增强底层网络潜在危险数据未激活状态挖掘的精度和效率。

Of the potential danger of underlying network inactive state data attributes with high fuzziness, traditional datamining methods based on attack behavior characteristics, potential hazards data inactive state of digging, the more prone tomining center, has the high false alarm rate, puts forward a dynamic mapping potentially dangerous data inactive state min？ing method, oriented to the underlying potential dangerous high dimension data attributes, by the method of principal com？ponent analysis, a potential danger to the underlying network data dimension reduction operation, the construction of poten？tially dangerous data mapping network, search the outstanding display neighborhood preserving network, generated fromthe network map, analysis of its parameters, obtain abnormal data, is inactive state data. Experimental results indicate thatthe proposed method can enhance the underlying network potential danger not active data mining efficiency and precision.