Stock Closing Price Interval Prediction Based on CEEMDAN-WTD-Bilstm-Transformer Model
DOI:
https://doi.org/10.54691/bcpbm.v20i.898Keywords:
BiLSTM; interval prediction; CEEMDAN; wavelet threshold de-noising; stock indices.Abstract
Stock price indices usually contain much noise and exhibit significant non-linearity, non-smoothness, and other complex characteristics. To address the complexity and long-term dependence of financial time series forecasting, a novel stock interval forecasting model based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise Analysis (CEEMDAN)-wavelet threshold de-noising (WTD) joint de-noising method and Transformer model improved by a Bi-directional Long Short-Term Memory model (BiLSTM) is proposed in this paper. First, based on CEEMDAN, the WTD method is incorporated to extract features from the basic ticker data of financial time series and use them as the input of the Transformer model. Meanwhile, the BiLSTM provides the time-series relationship among the input stock data for the Transformer to forecast the financial time series. Moreover, a hybrid sliding Kernel Density Interval Estimation (KDE) method to demonstrate the forecasting accuracy of financial time series. Using stock price index data, the prediction results are compared with those of traditional neural networks, and the results show that the stock interval prediction of the CEEMDAN-WTD-BiLSTM-Transformer model has a relatively high prediction accuracy, which achieves an improvement of 0.88%~3.62%.
Downloads
References
Cheng Zhaozhi. Research on financial time series forecasting based on deep learning. University of Electronic Science and Technology, 2021. doi: 10.27005/d.cnki.gdzku. 2021.004845.
Yang W, Han A, Wang Shouyang. Research on financial time series forecasting based on interval-type data. Journal of Systems Engineering,2016,31(06): 816-830.DOI: 10.13383/j.cnki.jse.2016.06.010.
Yan H, Ouyang H. Financial time series prediction based on deep learning. Wireless Personal Communications, 2018, 102(2): 683-700.
Hochreiter S, Schmidhuber J. Long short-term memory.Neural Computation, 1997, 9(8): 1735-1780.
Moghar A, Hamiche M. Stock Market Prediction Using LSTM Recurrent Neural Network. Procedia Computer Science, 2020, 170:1168-1173.
Peng Y, Liu YH, Zhang RF. Modeling and analysis of stock price prediction based on LSTM. Computer Engineering and Applications, 2019, 55(11):4.
Zhang Y, Chu G, Shen D. The Role of Investor Attention in Predicting Stock Prices: The Long Short-term Memory Networks Perspective. Finance Research Letters, 2020, 38(2):101484.
SuiJinCheng. Research on Stock Forecasting based on LSTM Neural Network. Qingdao: Qingdao University of Science and Technology, 2020.
He Yiyue, Li Ping, Han Jinbo. Research on stock market index forecasting modeling based on CEEMDAN⁃LSTM. Statistics and Information Forum, 2020, 35(6): 34 -45.
Yan Hongju. Deep learning-based integrated forecasting of financial time series data [J]. Statistics and Information Forum, 2020, 35(4): 33 -41.
Li Yafeng, Wang Hongbo, Li Chen, Wang Fuhao, Liu Meng, Luo Jingjing. An LSTM futures investment strategy incorporating attention mechanism. Computer System Applications,2021,30(08): 22-30.DOI: 10.15888/j.cnki.csa.008110.
Li Q, Liu N., Yan B. S., Song U. An improved EMD algorithm based on Cubic Hermite. Machine Tools and Hydraulics,2021,49(21):186-190.
TANG B, DONG S, SONG T. Method for eliminating mode mixing of empirical mode decomposition based on the revised blind source separation. Processing, 2012,92(1):248-258.
Xue F, Sun XW, Dong Zixue, Yang HR, Wang HY. Research on EEMD-based data noise reduction algorithm. Mechanical Engineering and Automation,2021(05):9-11.
ZHANG Jianwei,HOU Dove,TOU Zhenlei,ZHANG Yina,MA Ying. Noise reduction method for vibration signals of drainage structures based on CEEMDAN and SVD. Vibration and Shock, 2017,36(22): 138-143.DOI: 10.13465/j.cnki.jvs.2017.22.022.
Zhang Jianwen,Liu Yang,Zhang Dapeng,Zhang Huanyu. A new method of de-noising based on CEEMDAN with wavelet adaptive thresholding Electrical Measurement and Instrumentation,2018,55(10):14-18+33.
Bie Fengfeng,Zhang Shijia,Pei Junfeng,Xu Pengqing. Research on rolling bearing fault pattern recognition method based on wavelet thresholding and CEEMDAN joint de-noising. Mechanical Design and Manufacture,2020(02): 68-71.DOI: 10.19356/j.cnki.1001-3997.2020.02.017.
Zhao K,Zheng S-X. Research on noise reduction of wind turbine data based on improved EMD and wavelet thresholding method. Journal of Shanghai Electric Power University,2020,36(02):136-140.
Cai Cai, ZHAO Xiaotao, HU Xianeng, HUANG Xianghai, CHEN Huiming.Research on the combined CEEMDAN-wavelet thresholding method for ball mill cylinder vibration signal de-noising. Mechanical Science and Technology,2020,39(07): 1077-1085.DOI: 10.13433/j.cnki.1003-8728.20190241.
Zhu Jianjun,Zhang Zhetao,Kuang Cuilin,Pan Jiabao. A reliable wavelet de-noising quality evaluation index. Journal of Wuhan University (Information Science Edition),2015,40(05): 688-694.DOI:10.13203/j. whugis20130417.
Lu Tieding,Qian Wenlong,He Xiaoxing,Le Ying. An improved EMD method for determining the partitioned IMF components. Geodesy and Geodynamics,2020,40(07): 720-725.DOI: 10.14075/j.jgg.2020.07.012.
Volker D. Rhythm and Speech Rate: A Variation Coefficient for ΔC. Language and language-processing, 2006(1):231-241.
Vaswani, Ashish, et al. "Attention is all you need." Advances in neural information processing systems 30 (2017): 5998-6008.
Jielin Qiu and Jiacheng Zhu and Michael Rosenberg and Emerson Liu and Ding Zhao. Optimal Transport based Data Augmentation for Heart Disease Diagnosis and Prediction. arXiv preprint arXiv:2202.00567, 2022.
Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 770–778, 2016.
Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E Hinton. Layer normalization. arXiv preprint arXiv:1607.06450, 2016.
SuiJinCheng. Research on Stock Forecasting based on LSTM Neural Network Qingdao: Qingdao University of Science and Technology, 2020.
Yang XY, Zhang YF, Ye TZ, et al. Probabilistic interval prediction of wind power combination based on plain Bayesian. High Voltage Technology, 2020(3).
Tan, Meng Yu. Financial data prediction and classification based on support vector machine regression and learning. [Master's thesis]. Xi'an University of Electronic Science and Technology, 2014
T. Huang & L. Yu, Application of SDAE-LSTM model in financial time series forecasting. Computer Engineering and Applications, 2019. 55(01): pp. 142-148.
Kara Y, Boyacioglu M A, Baykan Ö K.Predicting direction of stock price index movement using artificial neural networks and support vector machines:The sample of the Istanbul Stock Exchange.Expert systems with Applications,2011,38(5):5311-5319.
Patel J,Shah S,Thakkar P,et al.Predicting stock and stock price index movement using trend deterministic data preparation and machine learning techniques.Expert Systems with Applications,2015,42(1):259-268.
Armano G,Marchesi M,Murru A. A hybrid genetic-neural architecture for stock indexes forecasting. Information Sciences,2005,170(1):3-33
Downloads
Published
Issue
Section
