×

Loading...
Ad by
  • 推荐 OXIO 加拿大高速网络，最低月费仅$40. 使用推荐码 RCR37MB 可获得一个月的免费服务
Ad by
  • 推荐 OXIO 加拿大高速网络，最低月费仅$40. 使用推荐码 RCR37MB 可获得一个月的免费服务

@

Here is my dumb answer.

yyhe(Jennifer)
Create an object for each customer corder. In this object, it has collection - items purchased.
Create a relationship table, has column productA, productB, relations(points).
Query customer orders data and build object order.
Go through each order and update the table by adding "1" to relations.

You should be able to find the strongest relationship. From this order history, you may find out customer's behavior.

In this relationship table, you may manually increase (give points to) some relationships.
You may aslo give more points if more than 1 are purchased.
(#2876484@0)
Last Updated: 2006-3-31
This post has been archived. It cannot be replied.
Report

Replies, comments and Discussions:

  • 工作学习 / 专业知识杂谈 / 求救--老板给的题目用SPSS算相关性。我统计早忘记了。 -kentridge(kentridge); 2006-3-15 {243} (#2840762@0)
    题目很简单。就是看客户订单中，如果一客户order某产品的话，他会不会也定另外的某种产品。我折腾了一夜。不会阿。特别是 mulit-dimensional corrolation，实在糊涂了。老板是个好人，他也知道难为我了。于是自己也在乱算，真不想拖他下水。帮帮忙，脑汁都干了。
    • To tell you a truth: Nothing is free in Canada. Give a price and somebody will help you out.. -hx001(hx001); 2006-3-15 (#2840966@0)
      • 其实都可以不用统计MODEL,直接用数据处理一下就可啦, 如果该客户没有其他产品订单, 当然他已经定的产品已这产品没有联系,另一办法就是用统计的MODEL,求出R的值, 如果R=0, 说明他们没有联系, R越大, 关系越大. 可惜我不会SPSS, 我只会SAS. -shelley57(shelley); 2006-4-6 (#2888540@0)
    • Is this related to Bussiness Intelligence or data warehousing? This is also about bussiness logic, I think it's same as upsell -simonqu(后悔); 2006-3-15 (#2841004@0)
    • A little free hint: Use SPSS data mining (Clementine)--Correlation, apriori... -tx101(numbereleven); 2006-3-15 (#2841358@0)
    • Here is my dumb answer. -yyhe(Jennifer); 2006-3-31 {578} (#2876484@0)
      Create an object for each customer corder. In this object, it has collection - items purchased.
      Create a relationship table, has column productA, productB, relations(points).
      Query customer orders data and build object order.
      Go through each order and update the table by adding "1" to relations.

      You should be able to find the strongest relationship. From this order history, you may find out customer's behavior.

      In this relationship table, you may manually increase (give points to) some relationships.
      You may aslo give more points if more than 1 are purchased.
    • Maybe it's too late to comment. But in your case, a simple cross table is good enough to answer your question. -wayne77(十年); 2006-4-6 (#2888460@0)

More Topics

  • For those of you, who don’t believe that we have technology to control the weather or cause earth quacks, please watch this video.
  • 美国考古博士生分析激光雷达数据发现失踪的千年玛雅古城
  • Yann LeCun最新公开表示：今年诺贝尔奖颁给AI，是诺奖委员会感到压力的结果，需要承认深度学习的影响。化学奖塞不下更多人，所以只能选了物理, 颁给Hinton和Hopefield勉强合理. 获奖成果玻尔兹曼机和Hopefield网络现在完全无用(completely useless)。
  • Nobel Prize-Winner on AI 'Existential Threat' 辛顿认为我们也许很快就失去对AI的控制。这东西在20年内可能超过人类。
  • CO2 is known as the gas of life according to this Harvard professor - the more CO2 we have the more life there can be on the planet.
    • 枫下论坛主坛 /  工作学习 /  专业知识杂谈