用于精准预测的人工蜂群聚类数据挖掘算法研究  

Research on Artificial Bee Colony Clustering Data Mining Algorithm for Accurate Prediction

作　　者：金红军[1] JIN Hong-jun(Yancheng Teacher's University,Yancheng Jiangsu 224002)

机构地区：[1]盐城师范学院,江苏盐城224002

出　　处：《数字技术与应用》2020年第10期95-97,共3页Digital Technology & Application

摘　　要：为了提高数据挖掘算法的查全率,为精准预测工作提供更加精准的数据支持,利用人工蜂群聚类技术在传统数据挖掘算法的基础上进行优化设计。针对不同的精准预测任务准备对应的数据样本,并通过选择、预处理和数据转换三个步骤,实现对初始样本数据的处理。利用人工蜂群聚类技术分类样本数据,并剔除离群数据。在设置关联规则的约束下,得出数据挖掘结果。通过算法性能的测试对比实验得出结论:与传统的数据挖掘算法相比,人工蜂群聚类数据挖掘算法的查全率提高了1.3%,将其应用到精准预测工作中,可以有效的降低预测误差。In order to improve the recall rate of data mining algorithm and provide more accurate data support for accurate prediction,artificial bee colony clustering technology is used to optimize the design based on the traditional data mining algorithm.The corresponding data samples are prepared for different precision prediction tasks,and the initial sample data is processed through three steps:selection,preprocessing and data conversion.Artificial bee colony clustering technology is used to classify sample data and eliminate outliers.Under the constraint of setting association rules,the results of data mining are obtained.Compared with the traditional data mining algorithm,the recall rate of the artificial bee colony clustering data mining algorithm is improved by 1.3%,which can effectively reduce the prediction error when it is applied to the accurate prediction.

关 键 词：精准预测 人工蜂群 聚类数据 数据挖掘算法 

分 类 号：TN929[电子电信—通信与信息系统]