A Hybrid Model Using Deep Learning to Predict Stock Price Index
Subject Areas : Computer Engineering
1 - Department of Accounting, Azadshahr Branch, Islamic Azad University, Azadshahr, Iran.
Keywords: stock price, Prediction, deep learning,
Abstract :
Predicting the stock price is a demanding task since multiple factors affect it. To enhance the stock price index prediction accuracy, the current study hybridizes variational mode decomposition (VMD) with the CNN-LSTM model. The proposed model, VMD-CNN-LSTM, works based on the decomposition-and-ensemble framework. To do this, VMD and CNN-LSTM were used to deal with the nonstationary and nonlinear nature of the stock price data. The former was first applied to the decomposition of time-series data into a number of components. Then, CNN-LSTM was applied to the prediction of the components. To end with, all the components’ prediction results were summed up to attain the final prediction result. To verify the effectiveness of the proposed model in terms of predicting the stock price index, its performance was compared to some single models as well as some VMD- and EMD-based hybrid models. The results not only confirmed the superiority of the hybrid models over the single ones, but also showed the higher effectiveness of VMD-based models compared to EMD-based ones regarding the prediction accuracy.
[1] Chhajer, P., Shah, M. and Kshirsagar, A., 2022. The applications of artificial neural networks, support vector machines, and long–short term memory for stock market prediction. Decision Analytics Journal, 2, p.100015.
[2] Gandhmal, D.P. and Kumar, K., 2019. Systematic analysis and review of stock market prediction techniques. Computer Science Review, 34, p.100190.
[3] Jakubowski, A. and Szewczak, Z.S., 2021. Truncated moments of perpetuities and a new central limit theorem for GARCH processes without Kesten’s regularity. Stochastic Processes and their Applications, 131, pp.151-171.
[4] Ding, Y., 2019, February. Empirical Analysis of Logarithmic Return Rate of China’s Financial Stocks—based on the ARMA-GARCH Model. In 2018 International Symposium on Social Science and Management Innovation (SSMI 2018) (pp. 286-290). Atlantis Press.
[5] Devi, B.U., Sundar, D. and Alli, P., 2013. An effective time series analysis for stock trend prediction using ARIMA model for nifty midcap-50. International Journal of Data Mining & Knowledge Management Process, 3(1), p.65.
[6] Gu, F.J. and Cen, Z.D., 2011. Study on the volatility of Chinese Shanghai and Shenzhen stock markets with GARCH and SV Models. J Math Pract Theory, 1, pp.4-22.
[7] Xu, X.S. and Chen, Y.B., 2001. Empirical study on nonlinearity in China stock market. Quantitative and Technical Economics, 18, pp.110-113.
[8] Liu, W. and Morley, B., 2009. Volatility forecasting in the hang seng index using the GARCH approach. Asia-Pacific Financial Markets, 16, pp.51-63.
[9] Yuan, S.L., Lu, H.M., Wang, P.P., Tian, C.G. and Gao, Z.J., 2021. Boron removal from metallurgical grade silicon by slag refining based on GA-BP neural network. Rare Metals, 40, pp.237-242.
[10] Aras, S. and Kocakoç, İ.D., 2016. A new model selection strategy in time series forecasting with artificial neural networks: IHTS. Neurocomputing, 174, pp.974-987.
[11] Raut, R.D., Priyadarshinee, P., Gardas, B.B. and Jha, M.K., 2018. Analyzing the factors influencing cloud computing adoption using three stage hybrid SEM-ANN-ISM (SEANIS) approach. Technological Forecasting and Social Change, 134, pp.98-123.
[12] Richhariya, B. and Tanveer, M., 2018. EEG signal classification using universum support vector machine. Expert Systems with Applications, 106, pp.169-182.
[13] Atmaja, B.T. and Akagi, M., 2021. Two-stage dimensional emotion recognition by fusing predictions of acoustic and text networks using SVM. Speech Communication, 126, pp.9-21.
[14] Fan, L., Pan, S., Li, Z. and Li, H., 2016. An ICA-based support vector regression scheme for forecasting crude oil prices. Technological Forecasting and Social Change, 112, pp.245-253.
[15] Zhang, G.P., 2003. Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing, 50, pp.159-175.
[16] Chen, Y. and Hao, Y., 2018. Integrating principle component analysis and weighted support vector machine for stock trading signals prediction. Neurocomputing, 321, pp.381-402.
[17] Ghosh, A., Bose, S., Maji, G., Debnath, N. and Sen, S., 2019, September. Stock price prediction using LSTM on Indian Share Market. In Proceedings of 32nd international conference on (Vol. 63, pp. 101-110).
[18] Baek, Y. and Kim, H.Y., 2018. ModAugNet: A new forecasting framework for stock market index value with an overfitting prevention LSTM module and a prediction LSTM module. Expert Systems with Applications, 113, pp.457-480.
[19] Wang, J. and Wang, J., 2017. Forecasting stochastic neural network based on financial empirical mode decomposition. Neural Networks, 90, pp.8-20.
[20] Bisoi, R., Dash, P.K. and Parida, A.K., 2019. Hybrid variational mode decomposition and evolutionary robust kernel extreme learning machine for stock price and movement prediction on daily basis. Applied Soft Computing, 74, pp.652-678.
[21] Yang, W., Peng, Z., Wei, K., Shi, P. and Tian, W., 2017. Superiorities of variational mode decomposition over empirical mode decomposition particularly in time–frequency feature extraction and wind turbine condition monitoring. IET Renewable Power Generation, 11(4), pp.443-452.
[22] Rawat, W. and Wang, Z., 2017. Deep convolutional neural networks for image classification: A comprehensive review. Neural computation, 29(9), pp.2352-2449.
[23] Krizhevsky, A., Sutskever, I. and Hinton, G.E., 2012. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25.
[24] Hochreiter, S. and Schmidhuber, J., 1997. Long short-term memory. Neural computation, 9(8), pp.1735-1780.