基于关联规则改进的网络异常数据挖掘方法

Network Anomaly Data Mining Method Based on Association Rule Improvement

传统的网络异常数据挖掘方法在计算网络异常数据与相关核心的距离时存在准确度不高的问题,导致挖掘精度有限,因此研究提出一种基于关联规则算法的改进网络异常数据挖掘方法。首先,初始化网络异常数据关联核心,采用Kmeans算法对网络异常实体数据执行局部搜索和优化;其次,运用关联规则算法精确计算网络异常数据与关联核心之间的距离;最后,确定距离关联核心最远的网络异常数据,以完成挖掘过程。研究结果显示,在挖掘相同数量的网络异常数据时,相较于传统方法,该研究方法能显著增加正确挖掘出的网络异常数据比例,提升对网络异常数据的识别精准度,具有显著优势。

Traditional methods for mining network anomaly data suffer from inaccuracies when calculating the distance between the network anomaly data and the associated centers,resulting in limited mining precision.So,this research proposes an improved network anomaly data mining method based on association rule algorithms.The method begins by initializing association centers for network anomaly data,followed by employing local search and optimization for network anomaly entities using the K-means algorithm;then,it accurately calculates the distance between network anomaly data and association centers using association rule algorithms;finally,it identifies the network anomalies that are furthest from the association centers to complete the mining process.Research results indicate that,when mining the same quantity of network anomaly data,this method significantly increases the proportion of accurately mined network anomalies compared to traditional methods,enhancing precise identification of network anomaly data,and having significant advantages.