Finansal Zaman Serileri Tahmininde Hibrit Yaklaşımlar: Bir Hisse Senedi Piyasası Uygulaması

Öz

Zaman serisi tahmininde hibrit yaklaşım, tekli modellerle karşılaştırıldığında en doğru modeli seçmede anahtar metodolojilerden biridir. Hisse senedi piyasası tahmini için hibrit modellemede makine öğrenmesi algoritmalarının uygulanması hızla gelişmektedir. Bu çalışmada, ikisi gelişmiş hisse senedi piyasasından (NASDAQ ve DAX) ve diğer ikisi yükselen hisse senedi piyasasından (NSE ve BIST) olmak üzere dört hisse senedi verisi için makine öğrenimi yaklaşımıyla hibrit modellemesi uygulanmıştır. Bir hisse senedi piyasası, değişken yapısıyla bilinir ve istikrarsız bir yapıya sahiptir, bu nedenle, bu çalışmada, en doğru zaman serisi tahmin modeline ulaşmak için oynaklığı dikkate alan çeşitli hibrit modeller önerilmektedir. Hibrid modellemede, öncelikle GARCH (Generalized Autoregressive Conditional Hetereoscedastic) ile birleştirilen ARIMA (Autoregressive Integrated Moving Average) modelleri zaman serilerinin modellemesinde, ardından SVM (support vector machine) ve LSTM (Long-Short term memory) gibi zeki modeller hata serilerinin doğrusal olmayan modellemesinde kullanılmaktadır. Ayrıca, hibrit modellerin performansları mevcut metodolijiler kullanılarak tekli modeller ile karşılaştırılmaktadır. Önerilen hibrit metodoloji, zaman serisi verisinin özelliklerini en iyi yansıtan birkaç modeli birleştirerek tahmin performanslarını önemli ölçüde iyileştirmektedir.

Anahtar Kelimeler

• Hibrit Yaklaşımlar
• Makine Öğrenimi
• Hisse Senedi Piyasası
• ARIMA, GARCH, SVM, LSTM

Hybrid Approaches in Financial Time Series Forecasting: A Stock Market Application

Öz

The hybrid approach in time series forecasting is one of the key methodologies in selecting the most accurate model when compared to the single models. Applications of machine learning algorithms in hybrid modeling for stock market forecasting have been developing rapidly. In this paper, we propose hybrid modeling through machine learning approach for four stock market data; two from the developed stock markets (NASDAQ and DAX) and the other two from the emerging stock markets (NSE and BIST). A stock market is known with its volatile structure and has an unstable nature, so we propose several combinations for the hybrid models considering volatility to reach the most accurate time series forecasting model. In hybrid modeling, first ARIMA (Autoregressive Integrated Moving Average) models combined with GARCH models (Generalized Autoregressive Conditional Heteroscedasticity) are used for modeling of time series, then intelligent models such as SVM (support vector machine) and LSTM (Long-Short term memory) are used for nonlinear modeling of error series. We also compare their performances with single models. The proposed hybrid methodology markedly improves the prediction performances of time series models by combining several models which reflect the time series data characteristics best.

Anahtar Kelimeler

• Hybrid time series
• machine learning
• ARIMA
• GARCH
• SVM
• LSTM
• stock market
• Hibrit Yaklaşımlar
• Makine Öğrenimi
• Hisse Senedi Piyasası

Kaynakça

1. Bildirici, M. & Ersin, Ö. Ö. (2009). Improving forecasts of GARCH family models with the artificial neural networks: An application to the daily returns in Istanbul Stock Exchange, Expert Systems with Applications, 36(4), 7355-7362. doi:10.1016/j.eswa.2008.09.051.
2. Bollerslev, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31(3), 307–327.
3. Box, G., & Jenkins, G.M., (1970). Time series analysis: forecasting and control. Holden-Day, San Francisco,CA.
4. Box, G.E., Jenkins, G.M., Reinsel, G.C., & Ljung, G.M. (2015), Time Series Analysis: Forecasting and Control, John Wiley and Sons.
5. Chen, K., Zhou, Y., Dai, F. (2015). A LSTM-based method for stock returns prediction: A case study of China stock market. In Proceedings of the 2015 IEEE international conference on big data (Big Data) IEEE, (pp. 2823–2824).
6. Chiang, W.C.,Urban, T.L., & Baildridge, G.(1996), A neural network approach to mutual fund net asset value forecasting, Omega 24 (2), 205–215.
7. de Mattos Neto, P.S., Cavalcanti, G.D., & Madeiro, F.(2017), Nonlinear combination method of forecasters applied to PM time series, Pattern Recognit. Lett. 95, 65–72.
8. Domingos S.O., Oliveira de J.F.L., & Mattos Neto de P.S.G., (2019), An intelligent hybridization of ARIMA with machine learning models for time series forecasting, Knowledge-Based Systems 175, pp.72-86

9. Engle, R. F. (1982). Autoregressive conditional heteroscedasticity with estimates of he variance of United Kingdom inflation. Econometrica: Journal of the Econometric Society, 987–1007.
10. Fayyad, R. Uthurusamy (Eds.), Proceedings of the First International Conference on Knowledge Discovery and Data Mining, AAAI Press, Menlo Park, CA.
11. Hyndman, Rob. 2006. “Another Look at Forecast Accuracy Metrics for Intermittent Demand.” Foresight: The International Journal of Applied Forecasting, 4, 43–46.
12. Khashei, M., & Bijari, M.(2010), An artificial neural network (p, d, q) model for time series forecasting, Expert Syst. Appl. 37(1), 479–489.
13. Khashei, M., Bijari, M.(2011), A novel hybridization of artificial neural networks and ARIMA models for time series forecasting, Appl. Soft Comput. 11(2), 2664–2675.
14. Kim K-J., (2003) Financial time series forecasting using support vector machines, Neurocomputing 55, pp.307 – 319.
15. Kim, H.Y., & Won, C.H. (2018), Forecasting the volatility of stock price index: A hybrid model integrating LSTM with multiple GARCH-type models, Expert Systems With Applications 103, 25–37.
16. Maknickiené N., & Maknickas, A. (2012, May). Application of neural network for orecasting of exchange rates and forex trading, In Proceedings of the 7th international scientific conference on business and management pp. 10–11.
17. Markham, L.S., & Rakes T.R. (1998), The effect of sample size and variability of data on the comparative performance of artificial neural networks and regression, Comput. Oper. Res. 25 251–263.
18. Muller, K.R., Smola, J.A., & Scholkopf, B.(1997), Prediction time series with support vector machines, Proceedings of International Conference on Artificial Neural Networks, Lausanne, Switzerland, pp. 999–1004.
19. Nelson, D. B. (1991). Conditional heteroskedasticity in asset returns: Anew approach. Econometrica: Journal of the Econometric Society, pp. 347–370.
20. Pérez-cruz, F., Afonso-rodríguez, J. A. & Giner, J. (2003), Estimating GARCH models using support vector machines, Quantitative Finance, 3(3), pp.163-172. doi:10.1088/1469-7688/3/3/302
21. Panigrahi, S., & Behera, H.(2017), A hybrid ETS–ANN model for time series forecasting, Eng.Appl. Artif. Intell. 66, 49–59.
22. Scholkopf, B., Burges, C., &Vapnik, V.(1995), Extracting support data for a given task, in: U.M. Sima, S.N., Neda, T., &Akbar, S.N. (2018), A Comparison of ARIMA and LSTM in Forecasting Time
23. Series,2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA) Vapnik, V.N.(1995), The Nature of Statistical Learning Theory, Springer, New York .
24. Yaser, S.A.M., & Atiya, A.F. (1996), Introduction to financial forecasting, Appl. Intell. 6, 205–213.
25. Zhang G.P., Patuwo E.B., & Hu M.Y., (1998) Forecasting with artificial neural networks: the state of the art, Int. J. Forecasting 14, pp.35–62.
26. Zhang, G.P. (2003), Time series forecasting using a hybrid ARIMA and neural network model, Neurocomputing, 50,159 – 175.