摘要
在电子商店经营管理中,利用孤立点方法发现购买总金额持续减少或增加以后,应该立即采取跟踪服务。本文提出了一种跟踪算法,该算法采用最小二乘法对“购买总金额-时间”进行线性拟合,求出斜率,根据斜率的大小判断购买的趋势。并通过实际数据的验证说明该算法的有效性。
In the management of Electronic Shopping,a trail service strategy is adopted when it is discovered that the total sum of money is decreasing or increasing continnously by using outliers.A new trail algorithm is provided in a linearity fitting of 'total money-time' by method of least squares.Thus,the purchasing tendency can be predicted according to the slope.Some numerical examples have been provided to illustrate the efficiency of the provided algorithm.
出处
《河南科技大学学报（自然科学版）》
CAS
2005年第4期41-43,共3页
Journal of Henan University of Science And Technology:Natural Science
基金
国家自然科学基金资助项目(60374023)
关键词
电子商店
数据挖掘
异常客户
跟踪服务
Electronic shopping
Data mining
Specific client
Trail service
