摘要
传统的异常数据监测算法依靠单台计算机对异常数据进行识别,识别速度慢,且无法满足对数据处理的精确性要求。针对上述问题,文中构建了Hadoop分布式财务异常数据分析模型。该模型采用Hadoop中的MapReduce框架作为并行计算框架,同时在数据异常检测算法方面引入了邻域关系的LOF算法,有效避免了数据集元素边缘可能会出现误判的情况。数值实验结果表明,文中所提算法的准确率相比其他3种同类算法提升了5%以上,且算法的总运行时间也明显缩短。由此可见,文中所提模型可快速、准确地检测出财务异常数据,保障医疗系统的平稳运行。
The traditional abnormal data monitoring algorithm relies on a single computer to identify the abnormal data, which is not only slow in recognition speed, but also can not meet the accuracy requirements of data processing. To solve the above problems, this paper constructs Hadoop distributed financial abnormal data analysis model. The model uses MapReduce framework in Hadoop as parallel computing framework, and introduces LOF algorithm of neighborhood relationship in data anomaly monitoring algorithm, which effectively avoids the possibility of misjudgment on the edge of data set elements. Numerical experiments show that the accuracy of the proposed algorithm is improved by more than 5% compared with the other three similar algorithms, and the total running time of the algorithm is also significantly shortened. Therefore, the proposed model can quickly and accurately detect the abnormal financial data, and ensure the smooth operation of the medical system.
作者
王金元
王宇
张亚松
林昊
龚致富
李盼
安新艳
WANG Jin-yuan;WANG Yu;ZHANG Ya-song;LIN Hao;GONG Zhi-fu;LI Pan;AN Xin-yan(The First Affiliated Hospital of Hebei North University,Zhangjiakou 075000,Hebei Province,China)
出处
《信息技术》
2022年第1期21-25,31,共6页
Information Technology
基金
河北省2017年度医学科学研究重点课题计划(20-170808)。
