基于机器学习的产品订单数据分析与未来需求精准预测研究  

Research on Product Order Data Analysis and Accurate Prediction of Future Demand Based on Machine Learning

作　　者：张钰婷 ZHANG Yuting(Dalian International Studies University,Dalian Liaoning 116000,China)

机构地区：[1]大连外国语大学,辽宁大连116000

出　　处：《信息与电脑》2024年第12期16-19,共4页Information & Computer

摘　　要：近年来,企业外部环境日益复杂多变,导致供应链面临诸多挑战。需求预测作为供应链管理的关键环节,其准确性直接影响企业的运营效率和市场竞争力。然而,传统需求预测方法受多种因素影响,预测准确率普遍较低。本文研究了基于机器学习算法的产品订单数据分析方法,旨在通过训练历史数据来精准预测未来需求。本文具体采用了长短时记忆网络(Long Short-Term Memory,LSTM)、随机森林、XGBoost和轻量梯度提升框架(Light Gradient Boosting Machine,LGBM)等模型,并对比了它们的预测性能。实验结果表明,LGBM模型在预测准确性上表现最优,能够为企业提供更加可靠的决策支持。In recent years,the external environment of enterprises has become increasingly complex and changing,leading to many challenges for the supply chain.As a key link in supply chain management,demand forecasting directly affects the operational efficiency and market competitiveness of enterprises.However,traditional demand forecasting methods are affected by various factors,resulting in generally low prediction accuracy.This article studies a product order data analysis method based on machine learning algorithms,aiming to accurately predict future demand by training historical data.This article specifically uses models such as LSTM,random forest,XGBoost,and LGBM,and compared their predictive performance.The experimental results show that the LGBM model performs the best in prediction accuracy and can provide more reliable decision support for enterprises.

关 键 词：机器学习 销量预测 深度学习 

分 类 号：F713.3[经济管理—产业经济]