基于机器学习的国际集装箱航运价格预测模型实证比较  

Empirical Comparison of International Container Shipping Price Prediction Models Based on Machine Learning

作　　者：焦岗汉 JIAo Gang-han(Zhejiang Seaport Asset Management Co.,Ltd.,Hangzhou 310005,China)

机构地区：[1]浙江海港资产管理有限公司,浙江杭州310005

出　　处：《浙江工商职业技术学院学报》2024年第2期15-19,共5页Journal of Zhejiang Business Technology Institute

摘　　要：针对国际集装箱航运价格波动问题,通过综合分析市场供求、运输成本等因素,构建了基于ARIMAX、LSTM和BP神经网络的预测模型。结果显示,LSTM模型在预测效果上表现最优,能有效记忆历史数据并准确预测未来趋势。这可为企业提供有效的市场分析工具和决策参考,有助于提升行业运营效率和风险管理水平。Addressing the volatility in international container shipping prices,prediction models based on ARI-MAX,LSTM,and BP neural networks are established through comprehensive analysis of factors such as market supply-demand dynamics and transportation costs.It is found that the LSTM model demonstrates the best predic-tive performance,effectively memorizing historical data and accurately forecasting future trends.This provides en-terprises with an effective market analytical tool and reference for decision-making,contributing to enhanced op-erational efficiency and risk management in the industry.

关 键 词：国际集装箱航运价格预测 模型研究 LSTM神经网络 

分 类 号：F553[经济管理—产业经济]