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
焦岗汉1
作者信息
- 1. 浙江海港资产管理有限公司,浙江 杭州 310005
- 折叠
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.
Abstract
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Key words
international container shipping price prediction/model research/LSTM neural network引用本文复制引用
出版年
2024
NETL
NSTL
万方数据