Abstract
In recent years, deep learning has been successfully applied to analyzing financial time series. In this paper, we propose a novel feature fusion learning (FFL) method to analyze the trend of high-frequency limit order books (LOBs). The proposed FFL method combines a convolutional neural network (CNN) and two long short-term memory (LSTM) models. The CNN module uses a kind of up-sampling techniques to enhance basic features and the two LSTM modules can extract time-related information from time-insensitive and time-sensitive features. In addition, two fusion rules (majority voting and weighted summation) are designed to fuse different feature models. Experiments are conducted on the benchmark dataset FI-2010. Experimental results show that FFL can go beyond the performance of every sub-model and outperform the state-of-the-art model on the prediction performance of LOBs.
This work was supported in part by the Natural Science Foundation of the Jiangsu Higher Education Institutions of China under Grant No. 19KJA550002, by the Six Talent Peak Project of Jiangsu Province of China under Grant No. XYDXX-054, by the Priority Academic Program Development of Jiangsu Higher Education Institutions, and by the Collaborative Innovation Center of Novel Software Technology and Industrialization.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Ntakaris, A., Magrisv, M., Kanniainen, J., et al.: Benchmark dataset for mid-price forecasting of limit order book data with machine learning methods. J. Forecast. 37, 852–866 (2018)
Tran, D.T., Magris, M., Kanniainen, J., et al.: Tensor representation in high-frequency financial data for price change prediction. In: 2017 IEEE Symposium Series on Computational Intelligence, pp. 1–7. IEEE (2017)
Kercheval, A.N., Zhang, Y.: Modelling high frequency limit order book dynamics with support vector machines. Quant. Finan. 15, 1315–1329 (2015)
Daiya, D., Wu, M., Lin, C.: Stock movement prediction that integrates heterogeneous data sources using dilated causal convolution networks with attention. In: 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 8359–8363 (2020)
Sirignano, J.A.: Deep learning for limit order books. Quant. Finan Intell. 19, 549–570 (2019)
Tsantekidis, A., Passalis, N., Tefas, A., et al.: Forecasting stock prices from the limit order book using convolutional neural networks. In: 2017 IEEE 19th Conference on Business Informatics (CBI), pp. 7–12 (2017)
Zhang, Z., Zohren, S., Roberts, S.: DeepLOB: deep convolutional neural networks for limit order books. IEEE Trans. Signal Process. 11, 3001–3012 (2019)
Tran, D.T., Iosifidis, A., Kanniainen, J., et al.: Temporal attention-augmented bilinear network for financial time-series data analysis. IEEE Trans. Neural Netw. Learn. Syst. 5, 1407–1418 (2018)
Tsantekidis, A., Passalis, N., Tefas, A., et al.: Using deep learning to detect price change indications in financial markets. In: 25th European Signal Processing Conference (EUSIPCO), pp. 2511–2515 (2017)
Kelotra, A., Pandey, P.: Stock market prediction using optimized Deep-ConvLSTM model. Big Data. 8, 5–24 (2020)
Ai, X.W., Hu, T., Bi, G.P., et al.: Discovery of jump breaks in joint volatility for volume and price of high-frequency trading data in China. In: International Conference on Knowledge Science, Engineering and Management, pp. 174–182 (2017)
Barbulescu, A., Bautu, E.: A hybrid approach for modeling financial time series. Int. Arab J. Inf. Technol. 9, 327–335 (2012)
Tokuoka, S., Yamawaki, M.T.: Trend predictions of tick-wise stock prices by means of technical indicators selected by genetic algorithm. Artif. Life Robot. 12, 180–183 (2008)
Zhou, X., Pan, Z., Hu, G., et al.: Stock market prediction on high-frequency data using generative adversarial nets. Math. Probl. Eng. 2018, 1–11 (2018)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9, 1735–1780 (1997)
Gers, F.A., Schraudolph, N.N., Schmidhuber, J.: Learning precise timing with LSTM recurrent networks. J. Mach. Learn. Res. 3, 115–143 (2002)
Chollet F.: Deep Learning with Python. Manning Publications Company, Shelter Island (2018)
Tran, D.T., Gabbouj, M., Iosifidis, A.: Multilinear class-specific discriminant analysis. Pattern Recogn. Lett. 100, 131–136 (2019)
Passalis N., Tsantekidis A., Tefas A., et al.: Time-series classification using neural bag-of-features. In: 2017 25th European Signal Processing Conference (EUSIPCO), pp. 301–305 (2017)
Passalis, N., Tefas, A., Kanniainen, J., et al.: Deep temporal logistic bag-of-features for forecasting high frequency limit order book time series. In: 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 7545–7549 (2019)
Tsantekidis, A., Passalis, N., Tefas, A., et al.: Using deep learning for price prediction by exploiting stationary limit order book features. Appl. Soft Comput. 93(106401), 1–10 (2020)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Lv, X., Zhang, L. (2021). Feature Fusion Learning Based on LSTM and CNN Networks for Trend Analysis of Limit Order Books. In: Mantoro, T., Lee, M., Ayu, M.A., Wong, K.W., Hidayanto, A.N. (eds) Neural Information Processing. ICONIP 2021. Lecture Notes in Computer Science(), vol 13111. Springer, Cham. https://doi.org/10.1007/978-3-030-92273-3_11
Download citation
DOI: https://doi.org/10.1007/978-3-030-92273-3_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-92272-6
Online ISBN: 978-3-030-92273-3
eBook Packages: Computer ScienceComputer Science (R0)