The solution using unsupervised learning (Kmeans, PCA, etc.) , combined with data visualization ,which had obtained effective and accurate classification of the corresponding customer, so that the solution can give the company an important reference for business decision-making
Meanwhile, the A/B test method is implemented to have a test of the effects for different customer classification on the company's business
近日,一家批发经销商尝试着针对一些客户改变其发货方式,从原来的每周五次每次早上发货,改为了更为便宜的每周三次每次晚上发货。起初,发货方式的改变并没有带来任何显著的负面结果,于是该批发商将这一更为便宜的变动推广到了所有客户。几乎同一时刻,该批发商开始收到客户对发货服务变动的投诉,也有的客户开始取消提货。该批发商受到的损失比节省下来的钱还要多。现在,该批发经销商雇佣你,希望你确定他们的客户特征和信息,以帮助它们在未来做出更加明智的商业决策。你的任务就是利用非监督学习技术,看看客户之间存在哪些相似之处,以及如何以最佳的方式将客户细分为不同类别。
This project requires Python 2.7 and the following Python libraries installed: