机器学习技术在期货市场运行预估中的应用

The Application of Machine Learning Technologies in the Prediction of Futures Market Running

通过分析期货品种历史运行规律,结合业务经验提取多类型数据特征,利用基于多个机器学习模型的融合技术,尝试预测未来五日期货合约交易量及持仓量走势。实证结果表明,该方法在一定程度上具备预估市场走势的能力,可以为业务人员提供帮助。

This paper analyzes the historical running pattern of different futures varieties,and applies a multi-model fusion technology to predict the market participation movement based on the features extracted from various sources.The experimental results show that the proposed model owns the ability of forecasting trading volume and holdings to some extent,and improve practical operational efficiency.