作者
陈敏敏
文章摘要
在加密货币高频交易的市场环境下,传统价格预测模型受限于数据高频波动特性与市场微观结构信息缺失等条件,难以实现精准预测。本研究提出一种融合人工智能(Artificial Intelligence,AI)与限价订单簿(Limited Orderbooks)数据的创新预测框架,通过深度学习算法挖掘历史交易数据中的潜藏的模式与趋势,同时将实时限价订单簿数据(包括买卖盘挂单分布、订单簿失衡率等关键指标)纳入模型,通过对投资者情绪与交易意图的量化表达,提升超短期价格预测精度。研究通过对比传统方法与新模型的预测效能,分析验证人工智能技术在金融量化分析领域的显著优势,以及订单簿数据对模型性能表现的重要贡献。研究结果为构建实时算法交易策略及完善动态风险管理机制提供了理论支撑与技术思路。通过量化微观结构信号与价格波动之间的因果关系,从而提升决策系统的时效性与准确性。
文章关键词
价格预测;限价订单簿;长短期记忆网络LSTM
参考文献
[1] Athanasia Dimitriadou and Andros Gregoriou.2023.Predicting Bitcoin Prices Using Machine Learning.Entropy 25,5(2023).
[2] Sepp Hochreiter and Jürgen Schmidhuber.1997.Long Short-Term Memory.Neural Computation 9,8(1997),1735-1780.
[3] Alexandra L’Heureux,Katarina Grolinger,Hany F.Elyamany,and Miriam A.M.Capretz.2017.Machine Learning With Big Data:Challenges and Approaches.IEEE Access 5(2017),7776-7797.
[4] Evelyn Olalekan-Elesin.2023.Comparing Machine Learning Models using Mean Absolute Error(MAE)Mean Absolute Percentage Error and(MAPE):A Comprehensive Analysis.Medium.
[5] Justin Sirignano and Rama Cont.2021.Universal features of price formation in financial markets:perspectives from deep learning.In Machine Learning and AI in Finance.Routledge,5-15.
[6] Cort J Willmott and Kenji Matsuura.2005.Advantages of the mean absolute error(MAE)over the root mean square error(RMSE)in assessing average model performance.Climate research 30,1(2005),79-82.
[7] Yuan Zhang.2013.Modeling High-Frequency Order Book Dynamics with Support Vector Machines.(2013).
Full Text:
DOI