摘要
传统供应商选择弊端较为明显,为此,提出基于大数据抓取的供应商不良行为数据监测方法。选取大数据抓取技术——网络爬虫技术获取供应商行为数据,按照时间先后顺序将供应商行为数据转换为平稳时间序列,利用拉格朗日插值算法补全行为数据的缺失,构建ARIMA模型,估计模型参数数值,确定不良行为数据辨识空间,通过拟合残差序列比较获取供应商不良行为数据辨识结果,实现不良行为数据的监测。通过实验数据可知,应用提出方法能够准确地辨识出多种类型的不良行为数据,证明提出方法具备更好的监测性能。
The disadvantages of traditional supplier selection are obvious.Therefore,the supplier bad behavior data monitoring method based on big data capture is proposed.We select big data capture technology,Web crawler technology,to obtain supplier behavior data,convert supplier behavior data into stable time series according to time sequence,use Lagrange interpolation algorithm to complete the lack of behavior data,build ARIMA model,estimate model parameter values,and determine the identification space of bad behavior data.The identification results of supplier's bad behavior data are obtained by fitting residual sequence comparison to realize the monitoring of bad behavior data.The experimental data show that the proposed method can accurately identify various types of bad behavior data,which proves that the proposed method has better monitoring performance.
作者
严峻
马路遥
龙铭
崔北为
YAN Jun;MA Luyao;LONG Ming;CUI Beiwei(Procurement Branch,State Grid Anhui Electric Power Co.,Ltd.,Hefei 230061,China)
出处
《微型电脑应用》
2023年第6期193-195,208,共4页
Microcomputer Applications