Enerji Sektörü Hisse Senedi Fiyat Tahminleri için Geliştirilmiş Sinir Ağlarında Ampirik Mod Ayrışımı: Petkim Petrokimya Holding A.Ş. Örneği

Öz

Finansal zaman serisi tahmini, yatırım stratejileri ve risk yönetimi uygulamalarında kritik bir rol oynamaktadır. Enerji sektörü hisselerinin karmaşık ve dinamik yapısı, hisse senedi fiyatlarının isabetli şekilde öngörülmesini güçleştirmektedir. Geleneksel tahmin yöntemleri, finansal piyasaların Etkin Piyasa Hipotezi ile de vurgulanan doğrusal olmayan ve çok yönlü dinamiklerini tam olarak yansıtmada yetersiz kalabilmektedir. Bu doğrultuda, bu çalışmada Ampirik Mod Ayrıştırma (EMD) tekniğinin ileri beslemeli sinir ağları ile bütünleştirilerek, enerji sektöründeki hisse senedi fiyat tahminlerinin doğruluk ve güvenilirliğinin artırılması hedeflenmiştir. Örnek olay incelemesi olarak Petkim Petrokimya Holding A.Ş. seçilmiş, Ocak 2020 ile Ekim 2023 arasını kapsayan döneme ait hisse senedi fiyat verileri, ham petrol fiyatları ve USD/TRY döviz kuru verileri birlikte değerlendirilmiştir. Analiz verileri Yahoo Finance veri tabanından sağlanmıştır. Veri ön işleme adımlarının ardından, EMD yöntemi kullanılarak hisse senedi fiyatları içsel mod fonksiyonlarına (IMF) ayrıştırılmıştır. Sinir ağı modelinin eğitiminde, hem orijinal zaman serisi verileri hem de EMD ile elde edilen IMF bileşenleri girdi olarak kullanılmıştır. Model, veri setinin %75’iyle eğitilmiş, kalan kısımlar ise test ve doğrulama için ayrılmıştır. Bulgular, EMD tabanlı içsel mod fonksiyonlarının sinir ağı modeline entegre edilmesinin, Petkim hisse senedi fiyat hareketlerinin yönü ve genel eğilimlerini tahmin etmede olumlu bir katkı sağlayabileceğine işaret etmektedir. Bu bağlamda çalışma, gelişmiş sinyal işleme tekniklerinin Türkiye gibi gelişmekte olan ve volatil bir piyasanın enerji sektöründe hisse senedi fiyatlarının öngörülebilirliğinin artırılmasına yönelik potansiyel sunduğunu öne sürmekte ve bu alandaki literatüre ampirik bir bakış açısı katmaktadır.

Anahtar Kelimeler

• Ampirik Mod Ayrışımı
• Sinir Ağları
• Hisse Senedi Fiyat Tahmini
• Finansal Tahmin
• Enerji Piyasası
• İçsel Mod Fonksiyonları

Empirical Mode Decomposition in Neural Networks Developed for Energy Shares Price Predictions: Petkim Petrochemical Holding Inc. Example

Abstract

Financial time series forecasting plays a critical role in investment strategies and risk management applications. The complex and dynamic structure of energy sector stocks makes it difficult to accurately predict stock prices. Traditional forecasting methods may fall short in fully reflecting the non-linear and multi-faceted dynamics of financial markets, as emphasised by the Efficient Market Hypothesis. In this context, this study aims to enhance the accuracy and reliability of stock price forecasts in the energy sector by integrating the Empirical Mode Decomposition (EMD) technique with feedforward neural networks. Petkim Petrochemical Holding A.Ş. was selected as a case study, and stock price data, crude oil prices, and USD/TRY exchange rate data for the period between January 2020 and October 2023 were evaluated together. The analysis data were obtained from the Yahoo Finance database. Following data preprocessing steps, stock prices were decomposed into intrinsic mode functions (IMFs) using the EMD method. Both the original time series data and the IMF components obtained via EMD were used as inputs in training the neural network model. The model was trained using 75% of the dataset, with the remaining portions allocated for testing and validation. The findings indicate that integrating EMD-based intrinsic mode functions into the neural network model may contribute positively to predicting the direction and general trends of Petkim stock price movements. In this context, the study suggests that advanced signal processing techniques offer potential for improving the predictability of stock prices in the energy sector of a developing and volatile market such as Turkey, thereby contributing an empirical perspective to the literature in this field.

Keywords

• Empirical Mode Decomposition
• Neural Networks
• Stock Price Prediction
• Financial Forecasting
• Energy Shares
• Intrinsic Mode Functions

Kaynakça

1. Aghabozorgi, S., Seyed, S., A., ve Ying, W. T. (2015). Time-series clustering – A decade review. Information Systems, 53, 16-38. https://doi.org/10.1016/J.IS.2015.04.007
2. Bachelier, L. (1900). Théorie de la spéculation. Annales Scientifiques de l’École Normale Supérieure, 17, 21-86. https://doi.org/10.24033/asens.476
3. Bao, W., Yue, J., ve Rao, Y. (2017). A deep learning framework for financial time series using stacked autoencoders and long-short term memory. Plos One, 12(7). https://doi.org/10.1371/JOURNAL.PONE.0180944
4. Barra, S., Carta, S. M., Corriga, A., Podda, A. S., ve Recupero, D. R. (2020). Deep learning and time series-to-image encoding for financial forecasting. IEEE/CAA Journal of Automatica Sinica, 7(3), 683-692. https://doi.org/10.1109/JAS.2020.1003132
5. Borovykh, A., Bohte, S., ve Oosterlee, C. W. (2017). Conditional time series forecasting with convolutional neural networks. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 10614 LNCS, 729-730. https://arxiv.org/abs/1703.04691v5
6. Cao, J., Li, Z., ve Li, J. (2019). Financial time series forecasting model based on CEEMDAN and LSTM. Physica A: Statistical Mechanics and its Applications, 519, 127-139. https://doi.org/10.1016/J.PHYSA.2018.11.061
7. Chacón, H. D., vd. (2020). Improving financial time series prediction accuracy using ensemble empirical mode decomposition and recurrent neural networks. IEEE Access, 8, 117133-45. https://doi.org/10.1109/ACCESS.2020.2996981
8. Chen, J. F., Chen, W. L., Huang, C. P., Huang, S. H., ve Chen, A. P. (2017). Financial time-series data analysis using deep convolutional neural networks [Bildiri]. 7th International Conference on Cloud Computing and Big Data, CCBD, Macau, China, 87-92. https://doi.org/10.1109/CCBD.2016.027

9. Chimmula, V. K. R., ve Zhang, L. (2020). Time series forecasting of COVID-19 transmission in Canada using LSTM networks. Chaos, Solitons & Fractals, 135. https://doi.org/10.1016/J.CHAOS.2020.109864
10. Daubechies, I., Lu, J., ve Wu, H. T. (2011). Synchrosqueezed wavelet transforms: An empirical mode decomposition-like tool. Applied and Computational Harmonic Analysis, 30(2), 243-261. https://doi.org/10.1016/J.ACHA.2010.08.002
11. Du, Y., Wang, J., Feng, W., Pan, S., Qin, T., Xu, R., ve Wang, C. (2021). AdaRNN: Adaptive learning and forecasting of time series. International Conference on Information and Knowledge Management, 402-411. https://doi.org/10.1145/3459637.3482315
12. Dudukcu, H. V., Taskiran, M., Cam Taskiran, Z. G., ve Yildirim, T. (2023). Temporal convolutional networks with RNN approach for chaotic time series prediction. Applied Soft Computing, 133. https://doi.org/10.1016/J.ASOC.2022.109945
13. Gamboa, J. C. B. (2017). Deep learning for time-series analysis. https://arxiv.org/abs/1701.01887v1
14. Huang, H., Lou, Y., Lei, L., Li, H. (2017). WNN prediction model of stock price with input dimensions reduced by rough set [Bildiri]. 2017 International Conference on Education, Economics and Management Research (ICEEMR 2017). 107-110. Atlantis Press. https://doi.org/10.2991/ICEEMR-17.2017.27
15. Jaramillo, J., Velasquez, J. D., ve Franco, C. J. (2017). Research in financial time series forecasting with SVM: Contributions from literature. IEEE Latin America Transactions, 15(1), 145-153. https://doi.org/10.1109/TLA.2017.7827918
16. Jin, Z., Jin, Y., ve Chen, Z. (2022). Empirical mode decomposition using deep learning model for financial market forecasting, PeerJ Computer Science, 8. https://doi.org/10.7717/peerj-cs.1076
17. Kara, Y., Acar Boyacioglu, M., ve Baykan, Ö. K. (2011). Predicting direction of stock price index movement using artificial neural networks and support vector machines: The sample of the Istanbul Stock Exchange. Expert Systems with Applications, 38(5), 5311-5319. https://doi.org/10.1016/J.ESWA.2010.10.027
18. Lahmiri, S. (2016). A variational mode decompoisition approach for analysis and forecasting of economic and financial time series. Expert Systems with Applications, 55, 268-273. https://doi.org/10.1016/J.ESWA.2016.02.025
19. Lim, B., Arık, S., Loeff, N., ve Pfister, T. (2021). Temporal fusion transformers for interpretable multi-horizon time series forecasting. International Journal of Forecasting, 37(4), 1748-1764. https://doi.org/10.1016/J.IJFORECAST.2021.03.012
20. Liu, H., Qi, L., ve Sun, M. (2022). Short-Term stock price prediction based on CAE-LSTM method. Wireless Communications and Mobile Computing. https://doi.org/10.1155/2022/4809632
21. Livieris, I. E., Pintelas, E., ve Pintelas, P. (2020). A CNN–LSTM model for gold price time-series forecasting. Neural Computing and Applications, 32(23), 17351-17360. https://doi.org/10.1007/s00521-020-04867-x
22. Mahalakshmi, G., Sridevi, S., ve Rajaram, S. (2016, 07-09 Ocak). A survey on forecasting of time series data [Bildiri]. 2016 International Conference on Computing Technologies and Intelligent Data Engineering, ICCTIDE Kovilpatti, India. https://doi.org/10.1109/ICCTIDE.2016.7725358
23. Mallikarjuna, M., ve Rao, R. P. (2019). Evaluation of forecasting methods from selected stock market returns. Financial Innovation, 5(1), 1-16. https://doi.org/10.1186/s40854-019-0157-x
24. Mansor, R., Zaini, B. J., ve Yusof, N. (2019). Prediction stock price movement using subsethood and weighted subsethood fuzzy time series models. AIP Conference Proceedings, 2138(1). https://doi.org/10.1063/1.5121123
25. Masini, R. P., Medeiros, M. C., & Mendes, E. F. (2023). Machine learning advances for time series forecasting. Journal of Economic Surveys, 37(1), 76-111. https://doi.org/10.1111/JOES.12429
26. Nova, A. J., Mim, Z. Q., Rowshan, S., Islam, Md. R. U., Nurullah, M., ve Biswas, M. (2021). Stock market prediction on high-frequency data using ANN. Asian Journal of Research in Computer Science, 10(3), 1-12. https://doi.org/10.9734/AJRCOS/2021/V10I230241
27. Oreshkin, B. N., Carpov, D., Chapados, N., ve Bengio, Y. (2019). N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. 8th International Conference on Learning Representations, ICLR 2020. https://arxiv.org/abs/1905.10437v4
28. Sagheer, A., ve Kotb, M. (2019). Time series forecasting of petroleum production using deep LSTM recurrent networks. Neurocomputing, 323, 203-213. https://doi.org/10.1016/J.NEUCOM.2018.09.082
29. Sezer, O. B., Gudelek, M. U., ve Ozbayoglu, A. M. (2020). Financial time series forecasting with deep learning : A systematic literature review: 2005–2019. Applied Soft Computing, 90. https://doi.org/10.1016/J.ASOC.2020.106181
30. Sezer, O. B., ve Ozbayoglu, A. M. (2018). Algorithmic financial trading with deep convolutional neural networks: Time series to image conversion approach. Applied Soft Computing, 70, 525-538. https://doi.org/10.1016/J.ASOC.2018.04.024
31. Siami-Namini, S., Tavakoli, N., & Siami Namin, A. (2018, 17-20 Aralık). A comparison of ARIMA and LSTM in forecasting time series [Bildiri]. 17th IEEE International Conference on Machine Learning and Applications, ICMLA 2018, Orlando, FL, USA, 1394-1401. https://doi.org/10.1109/ICMLA.2018.00227
32. Tealab, A. (2018). Time series forecasting using artificial neural networks methodologies: A systematic review. Future Computing and Informatics Journal, 3(2), 334-340. https://doi.org/10.1016/J.FCIJ.2018.10.003
33. Thomakos, D. D., ve Nikolopoulos, K. (2015). Forecasting multivariate time series with the theta method. Journal of Forecasting, 34(3), 220-229. https://doi.org/10.1002/FOR.2334
34. Valtierra-Rodriguez, M., Amezquita-Sanchez, J. P., Garcia-Perez, A., ve Camarena-Martinez, D. (2019). Complete ensemble empirical mode decomposition on FPGA for condition monitoring of broken bars in induction motors. Mathematics, 7(9), 783. https://doi.org/10.3390/MATH7090783
35. Wang, Y. (2017). Stock market forecasting with financial micro-blog based on sentiment and time series analysis. Journal of Shanghai Jiaotong University (Science), 22(2), 173-179. https://doi.org/10.1007/S12204-017-1818-4
36. Wu, N., Green, B., Ben, X., ve O’Banion, S. (2020). Deep Transformer Models for Time Series Forecasting: The Influenza Prevalence Case. https://arxiv.org/abs/2001.08317v1
37. Wu, Z., Pan, S., Long, G., Jiang, J., Chang, X., ve Zhang, C. (2020). Connecting the dots: Multivariate time series forecasting with graph neural networks. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 753-763. https://doi.org/10.1145/3394486.3403118
38. Xu, Y. (2024). Financial time series forecasting based on long-short term memory, complete ensemble empirical mode decomposition with adaptive noise and batch normalization. Science and Technology of Engineering, Chemistry and Environmental Protection, 1(9). https://doi.org/10.61173/rfsdha76
39. Zaheer, S., Anjum, N., Hussain, S., Algarni, A. D., Iqbal, J., Bourouis, S., ve Ullah, S. S. (2023). A Multi parameter forecasting for stock time series data using LSTM and deep learning model. Mathematics, 11(3), 590. https://doi.org/10.3390/MATH11030590