DAI Yurui,AN Junxiu,LI Zhuowei.CNN-BiLSTM Financial Market Trend Prediction based on Signal Decomposition and Noise Reduction[J].Journal of Chengdu University of Information Technology,2023,38(01):28-36.[doi:10.16836/j.cnki.jcuit.2023.01.005]
基于信号分解降噪的CNN-BiLSTM金融市场趋势预测
- Title:
- CNN-BiLSTM Financial Market Trend Prediction based on Signal Decomposition and Noise Reduction
- 文章编号:
- 2096-1618(2023)01-0028-09
- Keywords:
- financial time series prediction; discrete wavelet transform(DWT); variational modal decomposition(VMD); convolutional neural network; bidirectional long-short memory network
- 分类号:
- TP391.1
- 文献标志码:
- A
- 摘要:
- 随着金融时间序列数据日趋复杂,如何捕捉金融数据未来多天的趋势变化成了难题。针对该问题提出了基于信号分解降噪和注意力机制的CNN-BiLSTM金融市场趋势预测模型(attention-based DWT-VMD-CBiLSTM)。首先利用离散小波变换(DWT)对原始金融股指序列进行降噪处理,然后利用变分模态分解(VMD)对降噪后的数据进一步分解为若干子序列。再结合多元基本面特征,利用基于注意力机制的CBiLSTM网络模型对各子序列进行多步预测,最后将各预测结果相加得到最终结果,实现较为长期的趋势预测。为证明所提出的模型性能,在不同金融股指数据集上与不同模型进行了实验比较。结果表明,提出的模型预测精度优于其他方法,在平均绝对误差(MAE)和平均百分比误差(MAPE)上分别达到12.28、0.39和80.27、0.71,在可决系数(R2)和可释方差值(EVS)上达到72%、74%和79%、69%的拟合度。
- Abstract:
- With the increasing complexity of financial time series data, it becomes a challenge to capture the trend changes of financial data for multiple days in the future.To address this problem, a financial market trend prediction model(attention-based DWT-VMD-CBiLSTM)based on CNN+BiLSTM with signal decomposition noise reduction and added attention mechanism is proposed. Firstly, the original financial stock index series are noise-reduced by using discrete wavelet transform(DWT),and then the noise-reduced data are further decomposed into several sub-series by using variational modal decomposition(VMD).Based on the first two steps combined with multivariate features,a CNN+BiLSTM network model based on the attention mechanism is used to make multi-step predictions for each sub-series, and finally the prediction results are summed to obtain the final results to achieve longer-term trend prediction. To demonstrate the performance of the proposed model,experimental comparisons are conducted with different models on different financial stock index datasets.The results show that the prediction accuracy of the proposed model outperforms other methods,reaching 12.28、0.39 and80.27、0.71 in mean absolute error(MAE)and mean percentage error(MAPE),respectively,and 72%,74% and 79%,69% in the coefficient of resolvability(R2)and interpretable variance value(EVS)of the fit.
参考文献/References:
[1] 赵洪科.面向互联网金融微观对象的数据挖掘方法及应用研究[D].合肥:中国科学技术大学,2018.
[2] Rangel-Gonzalez J A,Frausto-Solis J,González-Barbosa J J,et al.Comparative study of ARIMA methods for forecasting time series of the mexican stock exchange[J].Studies in Computational Intelligence,2018,749:475-485.
[3] Chen Y,Hao Y.A Feature Weighted Support Vector Machine and K-Nearest Neighbor Algorithm for Stock Market Indices Prediction[J].Expert Systems with Applications,2017,80(SEP.):340-355.
[4] 伯毅.基于XGBoost模型的短期股票预测[D].哈尔滨:哈尔滨工业大学,2018.
[5] Prieto A,Prieto B,Ortigosa E M,et al.Neural networks: An overview of early research,current frameworks and new challenges[J].Neurocomputing,2016,214:242-268.
[6] Park,Frank,C,et al.Deep learning networks for stock market analysis and prediction:Methodology,data representations,and case studies[J].Expert Systems with Application,2017,83:187-205.
[7] Deng Y,Fan H,Wu S.A hybrid ARIMA-LSTM model optimized by BP in the forecast of outpatient visits[J].Journal of Ambient Intelligence and Humanized Computing,2020.
[8] 庄家懿,杨国华,郑豪丰,等.基于多模型融合的CNN-LSTM-XGBoost短期电力负荷预测方法[J].中国电力,2021,54(5):46-55.
[9] Niu Z,Yu Z,Tang W,et al.Wind power forecasting using attention-based gated recurrent unit network[J].Energy,2020,196:117081.
[10] 黄婕,张丰,杜震洪,等.基于RNN-CNN集成深度学习模型的PM_(2.5)小时浓度预测[J].浙江大学学报(理学版),2019,46(3):370-379.
[11] 杨青,王晨蔚.基于深度学习LSTM神经网络的全球股票指数预测研究[J].统计研究,2019,36(3):65-77.
[12] 程艳,尧磊波,张光河,等.基于注意力机制的多通道CNN和BiGRU的文本情感倾向性分析[J].计算机研究与发展,2020,57(12):2583-2595.
[13] 李卫疆,漆芳,余正涛.基于多通道特征和自注意力的情感分类方法[J].软件学报,2021,32(9):2783-2800.
[14] 林靖皓,秦亮曦,苏永秀,等.基于自注意力机制的双向门控循环单元和卷积神经网络的芒果产量预测[J].计算机应用,2020,40(S1):51-55.
[15] Lu W,Li J,Wang J,et al.A CNN-BiLSTM-AM method for stock price prediction[J].Neural Computing and Applications,2020,33(10):1-13.
[16] 周怡娜,路敬祎,张勇,等.VMD-SG-WT去噪法及其在混沌去噪中的应用[J].东北石油大学学报,2020,44(4):113-120.
[17] Xue Y J,Cao J X,Wang D X,et al.Application of the Variational-Mode Decomposition for Seismic Time-frequency Analysis[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2016,9(8):1-11.
[18] Zhang Y,Chen B,Pan G,et al.A novel hybrid model based on VMD-WT and PCA-BP-RBF neural network for short-term wind speed forecasting[J].Energy Conversion and Management,2019,195(SEP.):180-197.
[19] Niu H,Xu K,Wang W.A hybrid stock price index forecasting model based on variational mode decomposition and LSTM network[J].Applied Intelligence,2020,50.
[20] Liu Y,Yang C,Huang K,et al.Non-ferrous metals price forecasting based on variational mode decomposition and LSTM network-ScienceDirect[J].Knowledge-Based Systems,2020,188:105006.
[21] Li Y,Wang S,Wei Y,et al.A New Hybrid VMD-ICSS-BiGRU Approach for Gold Futures Price Forecasting and Algorithmic Trading[J]. IEEE Transactions on Computational Social Systems,2021(99):1-12.
[22] 黄擎,曾向阳.小波分解和改进CNN相融合的水声目标识别方法[J/OL].哈尔滨工程大学学报,2022(2):1-7[2021-11-23].http://kns.cnki.net/kcms/detail/23.1390.U.20211026.1112.002.html.
[23] 李晓寒,贾华丁,程雪,等.基于改进遗传算法和图神经网络的股市波动预测方法[J/OL].计算机应用:1-13[2021-11-22].http://kns.cnki.net/kcms/detail/51.1307.TP.20210816.1658.012.html.
[24] 王艺霏,于雷,滕飞,等.基于长-短时序特征融合的资源负载预测模型[J/OL].计算机应用:1-9[2021-11-21].http://kns.cnki.net/kcms/detail/51.1307.TP.20210705.2001.017.html.
[25] Wang F K,Mamo T,Cheng X B.Bi-directional long short-term memory recurrent neural network with attention for stack voltage degradation from proton exchange membrane fuel cells[J].Journal of Power Sources,2020,461:228170.
[26] Shi G,Qin C,Tao J,et al. A VMD-EWT-LSTM-based multi-step prediction approach for shield tunneling machine cutterhead torque[J].Knowledge-Based Systems,2021,228.
[27] Bahdanau D,Cho K,Bengio Y.Neural Machine Translation by Jointly Learning to Align and Translate[J].Computer Science,2014.
相似文献/References:
[1]牛梦晨,徐梓欣,陈旭唏,等.分级图像显示技术研究[J].成都信息工程大学学报,2023,38(02):142.[doi:10.16836/j.cnki.jcuit.2023.02.003]
NIU Mengchen,XU Zixin,CHAN Yukhee,et al.Research on Scalable Image Display Technology[J].Journal of Chengdu University of Information Technology,2023,38(01):142.[doi:10.16836/j.cnki.jcuit.2023.02.003]
备注/Memo
收稿日期:2022-04-22
基金项目:国家自然科学基金资助项目(71673032)