机器学习在购买意图方面的应用

Application of machine learning in purchase intention

顾客是否成功购买商品,不仅与商品本身有关,而且与顾客所处区域、类型和特殊节日有关。互联网时代,各大购物网站都有海量的顾客购买信息,因此可以通过顾客对网站的使用和操作信息,使用机器学习算法来预测顾客购买此类商品的意向。本文使用随机森林算法、SVM算法和朴素贝叶斯算法建立模型,并采用五折交叉验证的方法选出这3个可靠的模型,预测顾客在线购买的可能性,最终通过准确率、召回率、F1值、AUC对模型进行评估。实验结果表明:随机森林更适合于在线购买意图的预测。

Whether a customer successfully purchases a product is not only related to the product itself,but also related to the region,type and special festival where the customer is located.In the Internet era,all major shopping sites have a large amount of customer purchase information,so you can use machine learning algorithms to predict the customer's intention to purchase such products through the customer's use and operation of the website.In this paper,the random forest algorithm,SVM algorithm and Naive Bayes algorithm are used to establish the model,and the three reliable models are selected by the method of 5-fold crossValidation to predict the possibility of customers buying online.Finally,the accuracy rate,recall rate,The F1 value and AUC evaluate the model.The experimental results show that random forest is more suitable for online purchase intention prediction.