VAE-ATTGRU模型的股指期货价格预测研究

doi:10.3778/j.issn.1002-8331.2305-0465

摘要/Abstract

摘要： 针对股指期货市场高波动、非平稳、非线性和高信噪比等特性造成的预测难度大的问题，利用变分自编码器（VAE）和循环神经网络（RNN）提出一种基于VAE-ATTGRU的混合深度学习股指期货价格预测模型。利用变分自编码器对股指期货技术指标进行学习，将VAE学习到的潜在因子与原始数据融合实现数据增强，得到更丰富的因子表示；使用循环神经网络对股指期货价格进行预测，发现结合了注意力机制的门控循环单元（ATTGRU）可以对VAE增强后的股指期货数据进行充分学习，对关键特征信息进行捕捉并重新赋予权重。在沪深300股指期货、中证500股指期货和上证50股指期货数据上进行实验，通过均方根误差（RMSE）、平均绝对误差（MAE）、平均绝对百分比误差（MAPE）和决定系数R2对VAE-ATTGRU模型进行评估，发现其在预测精度上优于其他模型。

关键词: 股指期货预测, 变分自编码器（VAE）, 数据增强, 注意力机制, 门控循环单元（GRU）

Abstract: A hybrid deep learning stock index futures price prediction model based on VAE-ATTGRU is proposed, using variational autoencoder (VAE) and the recurrent neural network (RNN), to address the difficulty of predicting high-volatility, non-stationary, non-linear, and high signal-to-noise ratio characteristics in the stock index futures market. Firstly, the technical indicators of the stock index futures are learned using the VAE, and the latent factors learned by VAE are fused with the original data to achieve data augmentation and obtain a richer factor representation. Secondly, RNN is used to predict the stock index futures prices. It is found that the gated recurrent unit (GRU) combined with the attention mechanism (ATTGRU) can fully learn from the stock index futures data enhanced by VAE, capture key feature information, and reassign weights. The VAE-ATTGRU model is evaluated on datasets such as the CSI300 stock index futures, the CSI 500 index futures, and the SSE 50 index futures using root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination R2. The experimental results demonstrate that the VAE-ATTGRU model outperforms other models in terms of prediction accuracy.

Key words: stock index futures prediction, variational autoencoder (VAE), data augmentation, attention mechanism, gated recurrent unit (GRU)

张玉婷, 金传泰, 李勇. VAE-ATTGRU模型的股指期货价格预测研究[J]. 计算机工程与应用, 2024, 60(17): 293-301.

ZHANG Yuting, JIN Chuantai, LI Yong. Stock Index Futures Price Prediction Based on VAE-ATTGRU Model[J]. Computer Engineering and Applications, 2024, 60(17): 293-301.

参考文献

[1] 孙宏鑫, 魏先华. 基于趋势学习的混合神经网络股指期货预测研究[J]. 计量经济学报, 2021, 1(4): 921-934.
SUN H X, WEI X H. Research on stock index futures forecast based on trend learning and hybrid neural network[J]. China Journal of Econometrics, 2021, 1(4): 921-934.
[2] HUANG W, NAKAMORI Y, WANG S Y. Forecasting stock market movement direction with support vector machine[J]. Computers & Operations Research, 2005, 32(10): 2513-2522.
[3] 胡雪梅, 蒋慧凤. 具有技术指标的逻辑回归模型预测谷歌股票的涨跌趋势[J]. 系统科学与数学, 2021, 41(3): 802-823.
HU X M, JIANG H F. Logistic regression model with technical indicators predicts ups and downs for google stock prices[J]. Journal of Systems Science and Mathematical, 2021, 41(3): 802-823.
[4] BASAK S, KAR S, SAHA S, et al. Predicting the direction of stock market prices using tree-based classifiers[J]. The North American Journal of Economics and Finance, 2019, 47: 552-567.
[5] YU L, WANG S, LAI K K, et al. A multiscale neural network learning paradigm for financial crisis forecasting[J]. Neurocomputing, 2010, 73(4/5/6): 716-725.
[6] 黄卿, 谢合亮. 机器学习方法在股指期货预测中的应用研究——基于BP神经网络、SVM和XGBoost的比较分析[J]. 数学的实践与认识, 2018, 48(8): 297-307.
HUANG Q, XIE H L. Research on the application of machine learning in stock index futures forecast-comparison and analysis based on BP neural networks, SVM and XGBoost[J]. Mathematics in Practice and Theory, 2018, 48(8): 297-307.
[7] BUCCI A. Realized volatility forecasting with neural networks[J]. Journal of Financial Econometrics, 2020, 18(3): 502-531.
[8] LI S, TIAN Z, LI Y. Residual long short-term memory network with multi-source and multi-frequency information fusion: an application to China’s stock market[J]. Information Sciences, 2023, 622: 133-147.
[9] YAN W L. Stock index futures price prediction using feature selection and deep learning[J]. The North American Journal of Economics and Finance, 2023, 64: 101867.
[10] BARAK S, MIRAFZALI E, JOSHAGHANI M. Improving deep learning forecast using variational AutoEncoders[J]. arXiv:1906.01984, 2019.
[11] BOQUET G, MORELL A, SERRANO J, et al. A variational autoencoder solution for road traffic forecasting systems: missing data imputation, dimension reduction, model selection and anomaly detection[J]. Transportation Research Part C: Emerging Technologies, 2020, 115: 102622.
[12] NARITOMI Y, ADACHI T. Data augmentation of high frequency financial data using generative adversarial network [C]//Proceedings of the 2020 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, 2020: 641-648.
[13] CHEN W, SHI K. Multi-scale attention convolutional neural network for time series classification[J]. Neural Networks, 2021, 136: 126-140.
[14] GUNDUZ H. An efficient stock market prediction model using hybrid feature reduction method based on variational autoencoders and recursive feature elimination[J]. Financial Innovation, 2021, 7(1): 28.
[15] 邓德军, 徐洪珍, 韦诗玥. E-V-ALSTM模型的股价预测[J]. 计算机工程与应用, 2023, 59(6): 101-112.
DENG D J, XU H Z, WEI S Y. Stock price prediction based on E-V-ALSTM model[J]. Computer Engineering and Applications, 2023, 59(6): 101-112.
[16] KINGMA D P, WELLING M. An introduction to variational autoencoders[J]. Foundations and Trends? in Machine Learning, 2019, 12(4): 307-392.
[17] KINGMA D P, WELLING M. Auto-encoding variational bayes[J]. arXiv:1312.6114, 2013.
[18] GUNDUZ H, YASLAN Y, CATALTEPE Z. Intraday prediction of Borsa Istanbul using convolutional neural networks and feature correlations[J]. Knowledge-Based Systems, 2017, 137(1): 138-148.