基于深度学习的电子商务销售预测  

作　　者：王泽菡 徐毓晖 王晓文 

机构地区：[1]广州南方学院,广州510970 [2]广州数聚人工智能技术有限公司,广州511300

出　　处：《科技创新与应用》2024年第23期40-43,47,共5页Technology Innovation and Application

基　　金：广东大学生科技创新培育专项资金资助项目(pdjh2022b0642)。

摘　　要：随着手机端购物平台兴起,消费者购物行为出现明显改变。线上销售迅速成为各行各业主要销售渠道之一。传统的预测方法主要以数据挖掘为主,虽然简单易用,但是难以处理复杂的非线性时间序列,因此该文提出一种基于LSTM-DNN的预测模型,在同LSTM与RNN进行比较后,新模型具有明显优势,可有效提高预测精度,对电商企业降低管理成本具有重要意义。With the rise of mobile shopping platform,consumer shopping behavior has changed significantly.Online sales have quickly become one of the main sales channels in various industries.The traditional prediction method is mainly based on data mining,which is simple and easy to use,but it is difficult to deal with complex nonlinear time series,so this paper proposes a prediction model based on LSTM-DNN.Compared with LSTM and RNN,the new model has obvious advantages,effectively improves the prediction accuracy,and is of great significance for e-commerce enterprises to reduce management costs.

关 键 词：深度学习 电商销售 时间序列 购物平台 预测模型 

分 类 号：F713.36[经济管理—产业经济]