A hybrid deep learning approach by integrating LSTM-ANN networks with GARCH model for copper price volatility prediction
A hybrid deep learning approach by integrating LSTM-ANN networks with GARCH model for copper price volatility prediction
复制标题
将 LSTM-ANN 网络与 GARCH 模型相结合的混合深度学习方法用于铜价波动预测
DOI：
10.1016/j.physa.2020.124907
复制
发表时间：
2020-11-01
影响因子：
3.3
通讯作者：
Wen, Liu
中科院分区：
文献类型：
--
作者：
Hu, Yan;Ni, Jian;Wen, Liu
预测铜价波动是一项重要但具有挑战性的任务。鉴于影响铜价的众多因素的非线性和时变特征，我们提出了一种新颖的混合方法来预测铜价波动。该方法综合了两项重要技术。一种是经典的 GARCH 模型，它通过 GARCH 预测以紧凑的形式编码有关随时间变化的铜价波动性的有用统计信息。二是强大的深度神经网络，将GARCH预测与国内外市场因素相结合，寻找更好的非线性特征；它还将长短期记忆（LSTM）网络与传统的人工神经网络（ANN）相结合，以生成更好的波动性预测。我们的方法综合了这两种技术的优点，特别适合铜价波动预测的任务。实证结果表明，GARCH 预测可以作为信息特征显着提高神经网络模型的预测能力，而 LSTM 和 ANN 网络的集成是构建有用的深度神经网络结构以提高预测性能的有效方法。此外，我们对神经网络架构进行了一系列敏感性分析，以优化预测结果。结果表明，混合模型的 LSTM 和 BLSTM 网络之间的选择应考虑预测范围，而 ANN 配置应根据预测误差度量的选择进行微调。 (C) 2020 Elsevier B.V. 保留所有权利。
Forecasting the copper price volatility is an important yet challenging task. Given the nonlinear and time-varying characteristics of numerous factors affecting the copper price, we propose a novel hybrid method to forecast copper price volatility. Two important techniques are synthesized in this method. One is the classic GARCH model which encodes useful statistical information about the time-varying copper price volatility in a compact form via the GARCH forecasts. The other is the powerful deep neural network which combines the GARCH forecasts with both domestic and international market factors to search for better nonlinear features; it also combines the long short-term memory (LSTM) network with traditional artificial neural network (ANN) to generate better volatility forecasts. Our method synthesizes the merits of these two techniques and is especially suitable for the task of copper price volatility prediction. The empirical results show that the GARCH forecasts can serve as informative features to significantly increase the predictive power of the neural network model, and the integration of the LSTM and ANN networks is an effective approach to construct useful deep neural network structures to boost the prediction performance. Further, we conducted a series of sensitivity analyses of the neural network architecture to optimize the prediction results. The results suggest that the choice between LSTM and BLSTM networks for the hybrid model should consider the forecast horizon, while the ANN configurations should be fine-tuned depending on the choice of the measure of prediction errors. (C) 2020 Elsevier B.V. All rights reserved.