首页 > 分享 > 基于 GCN

基于 GCN

萌宠菠菠乐园
2025-05-19 07:22

浏览下载摘要关键词基金信息论文图表动态公开评议相关论文评论

基于 GCN-LSTM 模型的宠物产品销量预测

首发时间：2025-05-15

陈佳豪（2004-），男，学生，通信工程

刘宏娜1

刘宏娜（2005-），女，学术，计算机科学与技术

边廷玥1

边廷玥（1997-），女，讲师，软件无线电

1、辽宁工程技术大学，电子与信息工程学院，辽宁省兴城市，125105

摘要：随着宠物市场的迅速发展，准确预测宠物产品的销量对于企业的运营和决策至关重要。传统的销量预测方法往往难以充分捕捉宠物产品销售数据中的时空特征。本文提出了一种基于图卷积网络（GCN）和长短期记忆网络（LSTM）的模型，用于宠物产品销量的预测。该模型结合了 GCN 对空间特征的提取能力和 LSTM 对时间序列特征的处理能力，能够有效挖掘宠物产品销售数据中的时空依赖关系。通过对电商平台真实数据的实验验证，该模型在预测精度上优于传统的预测模型，为宠物产品的销售管理提供了更可靠的支持。

关键词：图卷积网络长短期记忆网络宠物产品销量预测时空特征

For information in English, please click here

Sales Forecasting of Pet Products Based on the GCN-LSTM Model

Chen Jiahao1

陈佳豪（2004-），男，学生，通信工程

Liu Hongna1

刘宏娜（2005-），女，学术，计算机科学与技术

Bian Tingyue1

边廷玥（1997-），女，讲师，软件无线电

1、Liaoning Technical University,College of Electronics and Information Engineering,Xingcheng City,Liaoning Province,125105

Abstract：With the rapid growth of the pet market, accurately predicting the sales of pet products is critical to the operation and decision-making of businesses. Traditional sales forecasting methods often struggle to fully capture the spatiotemporal characteristics of pet product sales data. In this paper, we propose a model based on Graph Convolutional Network (GCN) and Long Short-Term Memory Network (LSTM) for predicting the sales of pet products. The model combines the extraction ability of GCN for spatial features and the processing ability of LSTM for time series features, which can effectively mine the spatiotemporal dependencies in pet product sales data. Through the experimental verification of the real data of the e-commerce platform, the model is better than the traditional prediction model in terms of prediction accuracy, which provides more reliable support for the sales management of pet products.

Keywords： Graph Convolutional Networks Long Short-Term Memory Networks Sales Forecasting of Pet Products Spatiotemporal Features.

Click to fold

点击收起

基金：

论文图表：