NS-FEDformer模型对股票收益率的预测研究

doi:10.3778/j.issn.1002-8331.2404-0151

摘要/Abstract

摘要： 股票收益率预测作为金融市场中一项较为困难的预测任务，是研究者们重点关注的热门话题。近年来，Transformer类模型以其强大的特征表征能力和高效的并行计算能力，在序列建模中大放光彩。为了将深度学习技术更好地应用于股票预测领域，提出了基于NS-FEDformer模型的股票收益率预测方法，以在时间序列预测任务中表现优越的FEDformer模型为基础，引入non-stationary Transformer（NS-Transformer）框架中的de-stationary attention机制还原原始股票序列的注意力权重，提升了模型对于序列特征的提取能力。实验结果表明，NS-FEDformer模型在包括浙商证券、中国银河等十只股票数据集上的平均预测表现均优于深度学习股票预测的主流模型，且较经典的Transformer模型在不同预测步长下的MSE和MAE最大降低了30.35%和23.35%，RMSE和MAPE最大降低了16.65%和39.80%，验证了该模型的优越性。

关键词: 股票预测, 注意力机制, 深度学习, FEDformer模型

Abstract: As a difficult forecasting task in the financial market, stock return forecast is a hot topic that researchers focus on. In recent years, Transformer class model has gained prominence in sequence modeling with its powerful feature characterization capability and efficient parallel computing capability. In order to better apply deep learning technology to the field of stock prediction, a stock return prediction method based on NS-FEDformer model is proposed, which is based on FEDformer model with superior performance in time series prediction tasks. The de-stationary attention mechanism in the non-stationary Transformer (NS-Transformer) framework is introduced to restore the attention weight of the original stock series, which improves the ability of the model to extract sequence features. Experimental results show that the average prediction performance of NS-FEDformer model on ten stock datasets, including Zheshang Securities and China Galaxy, is better than that of mainstream deep learning stock prediction models, and compared with classic Transformer model, MSE and MAE decrease by 30.35% and 23.35%, RMSE and MAPE decrease by 16.65% and 39.80%, respectively, at the maximum under different prediction steps. The superiority of this model is verified.

Key words: stock forecast, attention mechanism, deep learning, FEDformer model

王婧, 李云霞. NS-FEDformer模型对股票收益率的预测研究[J]. 计算机工程与应用, 2025, 61(9): 334-342.

WANG Jing, LI Yunxia. Research on Stock Return Forecast by NS-FEDformer Model[J]. Computer Engineering and Applications, 2025, 61(9): 334-342.

参考文献

[1] ZHANG G P. Time series forecasting using a hybrid ARIMA and neural network model[J]. Neurocomputing, 2003, 50: 159-175.
[2] ENGLE R F. Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation[J]. Econometrica, 1982, 50(4): 987-1007.
[3] BOLLERSLEV T. Generalized autoregressive conditional heteroskedasticity[J]. Journal of Econometrics, 1986, 31(3): 307-327.
[4] COSTA-JUSSà M R, ALLAUZEN A, BARRAULT L, et al. Introduction to the special issue on deep learning approaches for machine translation[J]. Computer Speech & Language, 2017, 46: 367-373.
[5] FAYEK H M, LECH M, CAVEDON L. Evaluating deep learning architectures for speech emotion recognition[J]. Neural Networks, 2017, 92: 60-68.
[6] XING J L, LI K, HU W M, et al. Diagnosing deep learning models for high accuracy age estimation from a single image[J]. Pattern Recognition, 2017, 66: 106-116.
[7] TSANTEKIDIS A, PASSALIS N, TEFAS A, et al. Forecasting stock prices from the limit order book using convolutional neural networks[C]//Proceedings of the 2017 IEEE 19th Conference on Business Informatics. Piscataway: IEEE, 2017: 7-12.
[8] 潘水洋, 刘俊玮, 王一鸣. 基于神经网络的股票收益率预测研究[J]. 浙江大学学报(理学版), 2019, 46(5): 550-555.
PAN S Y, LIU J W, WANG Y M. Forecasting stock returns with artificial neural networks[J]. Journal of Zhejiang University (Science Edition), 2019, 46(5): 550-555.
[9] QIU J Y, WANG B, ZHOU C J. Forecasting stock prices with long-short term memory neural network based on attention mechanism[J]. PLoS One, 2020, 15(1): e0227222.
[10] 姚远, 张朝阳. 基于HP-LSTM模型的股指价格预测方法[J]. 计算机工程与应用, 2021, 57(24): 296-304.
YAO Y, ZHANG Z Y. Stock index price forecasting method based on HP filter[J]. Computer Engineering and Applications, 2021, 57(24): 296-304.
[11] 方义秋, 卢壮, 葛君伟. 联合RMSE损失LSTM-CNN模型的股价预测[J]. 计算机工程与应用, 2022, 58(9): 294-302.
FANG Y Q, LU Z, GE J W. Forecasting stock prices with combined RMSE loss LSTM-CNN model[J]. Computer Engineering and Applications, 2022, 58(9): 294-302.
[12] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems, 2017: 6000-6010.
[13] DING Q, WU S, SUN H, et al. Hierarchical multi-scale Gaussian Transformer for stock movement prediction[C]//Proceedings of the Twenty-Ninth International Conference on International Joint Conferences on Artificial Intelligence(IJCAI’20), 2020:4640-4646.
[14] 陈诗乐, 王笑, 周昌军. 基于GA-Transformer模型的多因子股票预测[J]. 广州大学学报(自然科学版), 2021, 20(1): 44-55.
CHEN S L, WANG X, ZHOU C J. Multi-factor stock forecasting method based on a GA-Transformer model[J]. Journal of Guangzhou University (Natural Science Edition), 2021, 20(1): 44-55.
[15] 任佳屹, 王爱银. 融合因果注意力Transformer模型的股价预测研究[J]. 计算机工程与应用, 2023, 59(13): 325-334.
REN J Y, WANG A Y. Causal attention Transformer model for stock price prediction[J]. Computer Engineering and Applications, 2023, 59(13): 325-334.
[16] ZHOU H Y, ZHANG S H, PENG J Q, et al. Informer: beyond efficient transformer for long sequence time-series forecasting[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2021: 11106-11115.
[17] LIU S, YU H, LIAO C, et al. Pyraformer: low-complexity pyramidal attention for long-range time series modeling and forecasting[C]//Proceedings of the 2022 International Conference on Learning Representations, Apr 25-29, 2022: 1-20.
[18] WU H, XU J, WANG J, et al. Autoformer: decomposition transformers with auto-correlation for long-term series forecasting[C]//Advances in Neural Information Processing Systems, 2021: 22419-22430.
[19] ZHOU T, MA Z Q, WEN Q S, et al. FEDformer: frequency enhanced decomposed transformer for long-term series forecasting[C]//Proceedings of the 39th International Conference on Machine Learning. Baltimore: PMLR, 2022: 27268-27286.
[20] PASSALIS N, TEFAS A, KANNIAINEN J, et al. Deep adaptive input normalization for time series forecasting[J]. IEEE Transactions on Neural Networks and Learning Systems, 2020, 31(9): 3760-3765.
[21] KIM T, KIM J, TAE Y, et al. Reversible instance normalization for accurate time-series forecasting against distribution shift[C]//Proceedings of the International Conference on Learning Representations, 2022.
[22] LIU Y, WU H X, WANG J M, et al. Non-stationary transformers: exploring the stationarity in time series forecasting[C]//Advances in Neural Information Processing Systems, 2022: 9881-9893.
[23] 陈东洋, 毛力. 融合增量学习与Transformer模型的股价预测研究[J]. 计算机科学与探索, 2024, 18(7): 1889-1899.
CHEN D Y, MAO L. Research on stock price prediction integrating incremental learning and Transformer model[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(7): 1889-1899.
[24] 林明松, 杨晓梅, 杨志霞. 结构化最大间隔双支持向量机在股票预测中的应用[J]. 计算机工程与应用, 2024, 60(11): 346-355.
LIN M S, YANG X M, YANG Z X. Structured maximum margin twin support vector machine and its application in stock trend prediction[J]. Computer Engineering and Applications, 2024, 60(11): 346-355.