Makine Öğrenimi ve Derin Öğrenme Algoritmalarını Kullanarak Hisse Senedi Fiyat Tahmini: Havacılık Sektörüne Yönelik Bir Örnek Çalışma
Abstract
Teknolojik ilerlemelerle birlikte, insanlar çeşitli elektronik cihazlar ve sensörler aracılığıyla sürekli olarak veri üretmekte ve bu veriler dijital ortamlarda depolanmaktadır. Bu büyük veri havuzu, yeni disiplinlerin doğmasına ve gelişmesine olanak tanıyan bir kaynak haline gelmiş; örneğin, veri bilimi, yapay zekâ, derin öğrenme ve nesnelerin interneti gibi alanlar ortaya çıkmıştır. Verilerin etkili bir şekilde yönetilmesi ve analiz edilmesi, modern işletmeler için rekabet avantajı sağlamaktadır. Bu çalışma, Borsa İstanbul'da (BIST) işlem gören Türk Hava Yolları AO (THYAO) şirketinin hisse senedi fiyatının tahmin edilmesini amaçlamaktadır. Bu amaçla, makine öğrenmesi algoritmalarından Support Vector Machine (SVM) ve Extreme Gradient Boosting (XGBoost) ile derin öğrenme algoritması olan Long Short-Term Memory (LSTM) kullanılmıştır. Modeller, 4 Ocak 2010 ile 5 Eylül 2023 tarihleri arasındaki günlük verileri içeren bir zaman diliminde eğitilmiştir. Gerçek hisse senedi fiyatları ile tahmin edilen fiyatlar karşılaştırılarak modellerin performansları değerlendirilmiş ve en düşük hataya sahip model belirlenmiştir. Önerilen modellerin performansları RMSE, MSE, MAE ve R2 hata istatistikleri kullanılarak değerlendirilmiştir. Elde edilen sonuçlara göre LSTM modelinin SVM ve XGBoost modellerine göre daha düşük hata katsayılarına sahip olduğu ve en iyi performansı verdiği belirlenmiştir.
Ethical Statement
Gerçekleştirilen bu çalışma bir saha çalışması olmadığından dolayı etik kurul onay sürecini gerektirmemektedir. Çalışmamız kamuya açık olan www.investing.com verilerine dayanmaktadır. Söz konusu verilerle makine öğrenmesi ve derin öğrenme algoritmaları kullanılarak tahminleme işlemi yapıldığı için de söz konusu süreçlere ihtiyaç doğmamıştır.
References
- İlkçar, M. (2023). Turkish Airlines BIST share price prediction with deep artificial neural network considering trading volume and seasonal values. International Journal of InformaticsTechnologies, 16(1), 43-53.
- Çınaroğlu, E, Avcı, T. (2020). Prediction of THY stock value with artificial neural networks. Atatürk University Journal of Economics and Administrative Sciences, 34(1), 1-19.
- Tokmak, M. (2022). Stock price prediction using Long-Short-term memory network. Mehmet Akif Ersoy University Journal of Applied Sciences, 6(2), 309-322.
- Fenghua, WEN, Jihong, XIAO, Zhifang, HE, Xu, GONG. (2014). Stock price prediction based on SSA and SVM. Procedia Computer Science, 31, 625-631.
- Pawar, K, Jalem, RS, Tiwari, V. (2019). Stock market price prediction using LSTM RNN. In Emerging Trends in Expert Applications and Security: Proceedings of ICETEAS 2018 (pp. 493-503). Springer Singapore.
- Yang, Y, Wu, Y, Wang, P, Jiali, X. (2021). Stock price prediction based on xgboost and lightgbm. In E3s web of conferences (Vol. 275, p. 01040). EDP Sciences.
- Kanakam, R, Ramesh, D, Mohmmad, S, Shabana, S, Prakash, TC. (2022, May). Stock price prediction using multiple linear regression and support vector machine (regression). In AIP Conference Proceedings (Vol. 2418, No. 1). AIP Publishing.
- Vuong, PH, Dat, TT, Mai, TK, Uyen, PH. (2022). Stock-price forecasting based on XGBoost and LSTM. Computer Systems Science & Engineering, 40(1).
- Kaneko, T, Asahi, Y. (2023). The Nikkei Stock Average Prediction by SVM. In International Conference on Human-Computer Interaction, 211-221.
- Gülmez, B. (2023). Stock price prediction with optimized deep LSTM network with artificial rabbits optimization algorithm. Expert Systems with Applications, 227, 120346.
- Almaafi, A, Bajaba, S, Alnori, F. (2023). Stock price prediction using ARIMA versus XGBoost models: the case of the largest telecommunication company in the Middle East. International Journal of Information Technology, 15(4), 1813-1818.
- Dezhkam, A, Manzuri, MT. (2023). Forecasting stock market for an efficient portfolio by combining XGBoost and Hilbert–Huang transform. Engineering Applications of Artificial Intelligence, 118, 105626.
- Schuster, M, Paliwal, K. (1997), Bidirectional recurrent neural networks. IEEE Trans. Signal Process. 1997, 45, 2673–2681.
- Hochreiter, S, Schmidhuber, J. (1997), Long Short-Term Memory. Neural Comput. 1997, 9, 1735–1780.
- Chen, X, Wei, L, Xu, J. (2017). House Price Prediction Using LSTM. http://arxiv.org/abs/1709.08432
- Chen, T, Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 13-17-Augu, 785–794. https://doi.org/10.1145/2939672.2939785
- Abar, H. (2020). Estimation of Gold Prices by Xgboost and Mars Methods. Ekev Academy Journal, (83), 427-446.
- Bakiler, H. (2023). Classification of gases with deep network based attributes and regression analysis of concentration values. Başkent University Institute of Science and Technology Unpublished Doctoral Thesis,2023
Stock Price Forecasting Using Machine Learning and Deep Learning Algorithms: A Case Study for the Aviation Industry
Abstract
With technological advances, humans are constantly generating data through various electronic devices and sensors, and this data is stored in digital environments. A vast amount of data has served as a valuable asset that has facilitated the rise and progression of novel fields, including data science, artificial intelligence (AI), deep learning (DL), and the internet of things (IoT). Effectively managing and analyzing data provides a competitive advantage for modern businesses. The objective of this study is to forecast the stock price of Turkish Airlines (THY), a publicly traded corporation listed on Borsa Istanbul. In order to achieve the intended objective, the utilization of machine learning approaches like SVM and XGBoost, as well as the deep learning algorithm Long Short-Term Memory (LSTM), are used. The models are trained over a time period including daily data from January 4, 2010 to September 5, 2023. The forecast performance of the models is evaluated by comparing the actual and predicted stock prices and the model with the lowest error is identified. The proposed models' performances are assessed using the RMSE, MSE, MAE, and R2 error statistics. According to the results obtained, it is determined that the LSTM model has lower error coefficients than SVM and XGBoost models and gives the best performance.
References
- İlkçar, M. (2023). Turkish Airlines BIST share price prediction with deep artificial neural network considering trading volume and seasonal values. International Journal of InformaticsTechnologies, 16(1), 43-53.
- Çınaroğlu, E, Avcı, T. (2020). Prediction of THY stock value with artificial neural networks. Atatürk University Journal of Economics and Administrative Sciences, 34(1), 1-19.
- Tokmak, M. (2022). Stock price prediction using Long-Short-term memory network. Mehmet Akif Ersoy University Journal of Applied Sciences, 6(2), 309-322.
- Fenghua, WEN, Jihong, XIAO, Zhifang, HE, Xu, GONG. (2014). Stock price prediction based on SSA and SVM. Procedia Computer Science, 31, 625-631.
- Pawar, K, Jalem, RS, Tiwari, V. (2019). Stock market price prediction using LSTM RNN. In Emerging Trends in Expert Applications and Security: Proceedings of ICETEAS 2018 (pp. 493-503). Springer Singapore.
- Yang, Y, Wu, Y, Wang, P, Jiali, X. (2021). Stock price prediction based on xgboost and lightgbm. In E3s web of conferences (Vol. 275, p. 01040). EDP Sciences.
- Kanakam, R, Ramesh, D, Mohmmad, S, Shabana, S, Prakash, TC. (2022, May). Stock price prediction using multiple linear regression and support vector machine (regression). In AIP Conference Proceedings (Vol. 2418, No. 1). AIP Publishing.
- Vuong, PH, Dat, TT, Mai, TK, Uyen, PH. (2022). Stock-price forecasting based on XGBoost and LSTM. Computer Systems Science & Engineering, 40(1).
- Kaneko, T, Asahi, Y. (2023). The Nikkei Stock Average Prediction by SVM. In International Conference on Human-Computer Interaction, 211-221.
- Gülmez, B. (2023). Stock price prediction with optimized deep LSTM network with artificial rabbits optimization algorithm. Expert Systems with Applications, 227, 120346.
- Almaafi, A, Bajaba, S, Alnori, F. (2023). Stock price prediction using ARIMA versus XGBoost models: the case of the largest telecommunication company in the Middle East. International Journal of Information Technology, 15(4), 1813-1818.
- Dezhkam, A, Manzuri, MT. (2023). Forecasting stock market for an efficient portfolio by combining XGBoost and Hilbert–Huang transform. Engineering Applications of Artificial Intelligence, 118, 105626.
- Schuster, M, Paliwal, K. (1997), Bidirectional recurrent neural networks. IEEE Trans. Signal Process. 1997, 45, 2673–2681.
- Hochreiter, S, Schmidhuber, J. (1997), Long Short-Term Memory. Neural Comput. 1997, 9, 1735–1780.
- Chen, X, Wei, L, Xu, J. (2017). House Price Prediction Using LSTM. http://arxiv.org/abs/1709.08432
- Chen, T, Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 13-17-Augu, 785–794. https://doi.org/10.1145/2939672.2939785
- Abar, H. (2020). Estimation of Gold Prices by Xgboost and Mars Methods. Ekev Academy Journal, (83), 427-446.
- Bakiler, H. (2023). Classification of gases with deep network based attributes and regression analysis of concentration values. Başkent University Institute of Science and Technology Unpublished Doctoral Thesis,2023
Details
| Primary Language | English |
|---|---|
| Subjects | Deep Learning |
| Journal Section | MBD |
| Authors | |
| Publication Date | March 28, 2024 |
| Submission Date | September 9, 2023 |
| Published in Issue | Year 2024 Volume: 36 Issue: 1 |
