Frequency Decomposition in Predictive Error Compensating Wavelet Neural Network

Abstract
_________________________

This paper presents an extended study of the previously proposed Predictive Error Compensation Neural Network (PECNET) model. Different frequencies are used as input, in addition with the use of the Butterworth filter and the model performances are compared. The results show that the PECNET with frequency decomposition and Butterworth filter applied to input data provides significantly more accurate predictions for stock price prediction problem with respect to previous studies and conventional machine learning and time series prediction methods without changing any hyperparameter or the structure. In addition, the time and space complexity of the PECNET model is less than all other compared machine learning methods.
Key Words: predictive error compensated neural network, Butterworth filter, frequency decomposition, wavelet transform, stock price forecasting

References
_________________________

1. Kindig, B. (2020), “New age of stock market volatility driven by machines”. Available at: https://www.forbes. com/sites/bethkindig/2020/04/10/new-age-of-stock-market- volatility-driven-by-machines/?sh=a53ca66dda21 (accessed 19 December 2022)
2. Fama, E.F. and French, K.R. (1993), “Common risk factors in the returns on stocks and bonds”. Journal of financial economics, Vol. 33 No. 1, pp.3-56. Available at: https://doi.org/10.1016/0304-405X(93)90023-5 (accessed 19 December 2022)
3. Gorenc, N.M. and Velušček, D. (2016), “Prediction of stock price movement based on daily high prices”. Quantitative Finance, Vol. 16 No.5, pp. 793-826. DOI: 10.1080/14697688.2015.1070960
4. Abu-Mostafa, Y.S. and Atiya, A.F. (1996), Introduction to financial forecasting. Applied intelligence, Vol. 6 No.3, pp.205-213. Available at: https://doi.org/10.1007/ BF00126626 (accessed 19 December 2022)
5. Box, G., (2013), “Box and Jenkins: time series analysis, forecasting and control”. In A Very British Affair. Palgrave Macmillan, London, pp.161-215.
6. Reddy, C.V. (2019), Predicting the stock market index using stochastic time series ARIMA modelling: the sample of BSE and NSE. Ind J Finance, Vol. 13 No.8, pp.7-25. Available at: http://dx.doi.org/10.2139/ssrn.3451677 (accessed 19 December 2022)
7. Khashei, M. and Bijari, M. (2010), An artificial neural network (p, d, q) model for timeseries forecasting. Expert Systems with applications, Vol. 37 No.1, pp.479-489. Available at: https://doi.org/10.1016/j. eswa.2009.05.044 (accessed 19 December 2022)
8. Hornik, K., Stinchcombe, M. and White, H. (1989), Multilayer feedforward networks are universal approximators. Neural networks, Vol. 2 No.5, pp.359- 366. Available at: https://doi.org/10.1016/0893- 6080(89)90020-8 (accessed 19 December 2022)
9. White, H. (1988), July. Economic prediction using neural networks: The case of IBM daily stock returns. ICNN Vol. 2, pp. 451-458. DOI: 10.1109/ ICNN.1988.23959.
10. Hassan, M.R., Nath, B. and Kirley, M. (2007), A fusion model of HMM, ANN and GA for stock market forecasting. Expert systems with Applications, Vol.33 No.1, pp.171-180. Available at: https://doi.org/10.1016/j. eswa.2006.04.007 (accessed 19 December 2022) 11. Wang, J.J., Wang, J.Z., Zhang, Z.G. and Guo, S.P. (2012), Stock index forecasting based on a hybrid model. Omega, Vol. 40 No.6, pp.758-766. Available at: https://doi.org/10.1016/j.omega.2011.07.008 (accessed 19 December 2022)
12. Hajizadeh, E., Seifi, A., Zarandi, M.F. and Turksen, I.B. (2012), A hybrid modeling approach for forecasting the volatility of S&P 500 index return. Expert Systems with Applications, Vol. 39 No.1, pp.431-436. Available at: https://doi.org/10.1016/j.eswa.2011.07.033 (accessed 19 December 2022)
13. Shahi, T.B., Shrestha, A., Neupane, A. and Guo, W. (2020), Stock price forecasting with deep learning: A comparative study. Mathematics, Vol. 8 No.9, p.1441. Available at: https://doi.org/10.3390/math8091441 (accessed 19 December 2022)
14. Bao, W., Yue, J. and Rao, Y. (2017), A deep learning framework for financial time series using stacked autoencoders and long-short term memory. PloS one, Vol.12 No.7, p.e0180944. Available at: https://doi. org/10.1371/journal.pone.0180944 (accessed 19 December 2022)
15. Li, J., Bu, H. and Wu, J. (2017), June. Sentimentaware stock market prediction: A deep learning method. 2017 international conference on service systems and service management. IEEE, pp. 1-6. DOI: 10.1109/ICSSSM.2017.7996306.
16. Goodfellow, I., Bengio, Y. and Courville, A. (2016), Deep learning. MIT press.
17. Ustundag, B.B. and Kulaglic, A. (2020), Highperformance time series prediction with predictive error compensated wavelet neural networks. IEEE Access, Vol. 8, pp. 210532-210541. DOI: 10.1109/ ACCESS.2020.3038724
18. Kulaglic, A. and Ustundag, B.B. (2021), “Stock price prediction using predictive error compensation wavelet neural networks”. Comput. Mater. Continua, Vol. 68 No.3, pp. 3577-3593. DOI:10.32604/ cmc.2021.014768.
19. Kulaglic, A. and Ustundag, B.B. (2021), “Predictive Error Compensating Wavelet Neural Network Model for Multivariable Time Series Prediction”. TEM Journal, Vol.10 No.4. DOI: 10.18421/TEM104‐61.
20. Ellis, G. (2012), Control system design guide: using your computer to understand and diagnose feedback controllers. Butterworth-Heinemann.
21. Moshiri, S. and Cameron, N. (2000), Neural network versus econometric models in forecasting inflation. Journal of forecasting, Vol.19 No.3, pp.201-217.
22. Patterson, D.W. (1996), Artificial Neural Networks: Theory and Applications, Prentice Hall.

Corresponding Author
Ajla Kulaglic,
Istanbul Technical University,
Istanbul,
Turkey,
E-mail: kulaglic@itu.edu.tr