基于决策树分类算法的电力营销数据异常识别方法  

Exception Identification Method of Power Marketing Data Based on Decision Tree Classification Algorithm

作　　者：曹斌 CAO Bin(State Grid Hubei Electric Power Co.,Ltd.Jiangling County Power Supply Company,Hubei Jiangling 434100 China)

机构地区：[1]国网湖北省电力有限公司江陵县供电公司,湖北江陵434100

出　　处：《长江信息通信》2024年第9期99-101,共3页Changjiang Information & Communications

摘　　要：针对现有识别方法在对电力营销数据异常识别时,存在识别精度较低,无法满足电力营销管理需要的问题,引入决策树分类算法,开展电力营销数据异常识别方法研究。采集电力营销数据,得到数据分层特征信息融合结果,计算数据异常识别信任度。结合决策树分类算法,构建异常识别模型。计算异常数据数据分布关联系数,将电力营销数据异常识别输出。通过对比实验证明,新的识别方法得到的识别结果精度更高,更具有实际应用价值。In view of the existing identification methods in the abnormal identification of power marketing data,there is a low identification accuracy,can not mect the nceds of power marketing management,the decision tree classification algorithm is introduced to carry out the research on the abnormal identification method of power markcting data.Collect the power marketing data,get the data hierarchical feature information fusion results,and calculate the trust degree of data abnormal identification.Combined with the decision tree classification algorithm.Calculate the correlation coefficient of abnormal data and data distribution,and identify the abnormal output of power marketing data.Through comparative experiments,it is proved that the new identification method has higher accuracy and more practical application value.

关 键 词：决策树分类算法 数据 识别 异常 电力营销 

分 类 号：TM731[电气工程—电力系统及自动化]