基于模糊聚类的电力通信网络信息异常值挖掘方法研究

Research on Information Outlier Mining Method for Power Communication Network Based on Fuzzy Clustering

由于电力通信网络信息异常值的隐蔽性、多样性和动态变化性,导致在挖掘过程中难以全面地识别出所有的异常值。因此,该文提出了一种基于模糊聚类的电力通信网络信息异常值挖掘方法。采用该方案可对电力通信网络信息进行数据填补缺失值、数据标准化处理与统一表示,构建一个包含异常值相关属性的集合矩阵,使用Apriori算法识别出异常特征之间的强关联规则,通过规则提取出异常值特征参数,基于模糊聚类算法,计算样本与聚类中心的相似度并设置隶属度阈值,挖掘出电力通信网络中的异常值。实验结果表明,该方法能够更有效地根据异常值的特性,挖掘出更多的异常值,为保障电力通信网络的安全提供有力的技术支撑。

Due to the concealment,diversity,and dynamic variability of information outliers in power communication networks,it is difficult to comprehensively identify all outliers during the mining process.Therefore,this article proposes a method for mining information outliers in power communication networks based on fuzzy clustering.Fill in missing values,standardize data,and unify representation of power communication network information,construct a set matrix containing outlier related attributes,use Apriori algorithm to identify strong association rules between outlier features,extract outlier feature parameters through rules,calculate the similarity between samples and cluster centers based on fuzzy clustering algorithm,and set membership threshold to mine outliers in the power communication network.The experimental results show that this method can more effectively deal with the characteristics of outliers,thereby mining more outliers and providing strong technical support for ensuring the security of power communication networks.