TÜRKİYE'DE DOLAR/TL KURUNU TAHMİN ETMEK: UZUN-KISA BELLEK SİNİR AĞLARI YAKLAŞIMI

Year 2023, Volume: 10 Issue: 2, 935 - 949, 02.08.2023

Ayten Yağmur , Zeynep Karaçor , Fatih Mangır , Abdul-razak Bawa Yussif

https://doi.org/10.30798/makuiibf.1097568

Abstract

Döviz kuru zaman serisinin tahmini oldukça zorlu, ancak önemli bir süreçtir. Bu, serilerdeki kalıtsal gürültü özelliğinin ve kırılgan davranışının sonucudur. Bu amaçla ARIMA gibi zaman serisi analiz modelleri kullanılmıştır. Ancak bu modeller döviz kurlarının stokastik özelliklerinin yanı sıra doğrusal olmama özelliklerini de açıklayamamaları nedeniyle sınırlıdırlar. Daha doğru bir döviz kuru tahmini gerçekleştirmek için, önemli başarı oranlarına sahip derin öğrenme yöntemleri uygulanmaktadır. Bu çalışma da, Türkiye'deki USD/TL kurunu tahmin etmek için Uzun-Kısa Vadeli Bellek Sinir Ağı yöntemi uygulanmaktadır. Bu makaleden elde edilen sonuç, Uzun-Kısa Süreli Bellek Sinir Ağı derin öğrenme yönteminin otoregresif hareketli ortalamalar yöntemi ile Çok katmanlı Yapay Sinir Ağı modellerine kıyasla daha yüksek tahmin yapmaktadır.

Keywords

Tahmin, Döviz Kuru, Zaman Serileri, ARIMA, LSTM, MLP.

PREDICTING USD/ TL EXCHANGE RATE IN TURKEY: THE LONG-SHORT TERM MEMORY APPROACH

Abstract

The prediction of the exchange rate time series has been quite challenging but is an essential process. This is as a result of the inherent noise and the volatile behavior in these series. Time series analysis models such as ARIMA have been used for this purpose. However, these models are limited due to the fact that they are not able to explain the non-linearity as well as the stochastic properties of foreign exchange rates. In order to perform a more accurate exchange rate prediction, deep-learning methods have been employed withremarkable rates of success. In this paper, we apply the Long-Short Term Memory Neural Network to predict the USD/TL exchange rate in Turkey. The result from this paper indicates that the Long-Short Term Memory Neural Network deep learning method gives higher prediction accuracy compared to the Auto Regressive Integrated Moving Average and the Multilayer Perception Neural Network models.

Keywords

Prediction, Exchange Rate, Time Series, ARIMA, LSTM, MLP.

References

• Azzouni, A. and Pujolle, G. (2017), “A Long Short-Term Memory Recurrent Neural Network Framework For Network Traffic Matrix Prediction”, arXiv preprint arXiv:1705.05690 (Accessed: 23.02.2021).
• Bengio, Y., Simard, P. and Frasconi, P. (1994), “Learning Long-Term Dependencies With Gradient Descent İs Difficult”, Ieee Transactions On Neural Networks, Vol. 5, No. 2: 157-166.
• Chandwani, D., and Manminder S.S. (2014), “Stock Direction Forecasting Techniques: An Empirical Study Combining Machine Learning System With Market İndicators İn The Indian Context”, International Journal of Computer Applications, Vol. 92, No. 11: 8-17.
• Das, S.R., Mishra, D. and Rout, M. (2020), “A Hybridized ELM-Jaya Forecasting Model For Currency Exchange Prediction”, Journal of King Saud University-Computer and Information Sciences, Vol. 32, No:3: 345-366.
• Faraway, J., and Chatfield, C. (1998), “Time Series Forecasting With Neural Networks: A Comparative Study Using The Airline Data”, Journal of the Royal Statistical Society: Series C (Applied Statistics), Vol. 47, No. 2: 231-250.
• Gers, F.A., Schmidhuber, J. and Cummins, F. (2000), “Learning To Forget: Continual Prediction With LSTM”, Neural Computation, Vol.12, No. 10: 2451-2471.
• Gers, F. A., Schraudolph, N.N. and Schmidhuber. J. (2002), “Learning Precise Timing With LSTM Recurrent Networks”, Journal Of Machine Learning Research, Vol. 3, No.1: 115-143.
• Graves, A., Jaitly, N. and Mohamed, A. (2013), “Hybrid Speech Recognition With Deep Bidirectional LSTM”, IEEE Workshop On Automatic Speech Recognition and Understanding, https://ieeexplore.ieee.org/document/6707742, (Accessed: 01.12.2020).
• Greff, K., Srivastava, R.K., Koutník,J., Steunebrink, B.R. and Schmidhuber. J. (2016),”LSTM: A search space odyssey”, IEEE transactions on neural networks and learning systems, Vol.28, No.10: 2222-2232.
• Hochreiter, S., Bengio, Y., Frasconi, P. and Schmidhuber. J. (2001),” Gradient flow in recurrent nets: the difficulty of learning long-term dependencies”, http://www.bioinf.jku.at/publications/older/ch7.pdf, (Accessed: 08.11.2020).
• Hu, Jiaojiao, Wang, Xiaofeng, Zhang, Ying, Zhang, Depeng, Zhang, Meng and Xue, Jianru (2020). “Time Series Prediction Method Based on Variant LSTM Recurrent Neural Network”, Neural Processing Letters, Vol. 5: 1–2
• Jaeger, H. (2002), “Tutorial On Training Recurrent Neural Networks, Covering BPPT, RTRL, EKF And The And The"Echo State Network" Approach”, https://citeseerx. ist.psu. edu/viewdoc/download? doi=10.1.1.378.4095&rep=rep1&type=pdf, (Accessed: 07.10.2021).
• Khairalla, M., and AL-Jallad. N.T. (2017), “Hybrid Forecasting Scheme For Financial Time-Series Data Using Neural Network And Statistical Methods”, International Journal of Advanced Computer Science and Applications, Vol.8, No.9: 319-327.
• Kleiner, B. (1977), "Time series analysis: Forecasting and control.", Technometrics, Vol. 19, No.3:343-344.
• Kong, Y., Huang, Q.. Wang, C., Chen, J., Chen, J. and He, D. (2018), “Long short-term memory neural networks for online disturbance detection in satellite image time series”, Remote Sensing Vol.10, No.3: 452.
• LeCun,Y., Bengio, Y. and Hinton, G. (2015), “Deep Learning”, Nature, Vol.521, No.7553: 436-444.
• Li, Y. and Cao, H. (2018), “Prediction For Tourism Flow Based On LSTM Neural Network”, Procedia Computer Science, Vol.129: 277-283.
• Okasha, M.K. (2014), “Using Support Vector Machines İn Financial Time Series Forecasting”, International Journal of Statistics and Applications, Vol.4, No.1: 28-39.
• Osório, G. J., Lotfi, Shafie-khah, M., Campos, V. and Catalão, J. (2019),”Hybrid Forecasting Model For Short-Term Electricity Market Prices With Renewable Integration”, Sustainability Vol.11, No.1: 57.
• Pallabi, P. and Kumari, B. (2017), ”Stock Market Prediction Using ANN SVM ELM: A Review”, International Journal of Emerging Trends & Technology in Computer Science, Vol.6, No.3: 88-94.
• Qiu, J., Wang, B., & Zhou, C. (2020). “Forecasting stock prices with long-short term memory neural network based on attention mechanism”, PloS one, Vol. 15, No.1:1-15
• Sujin, P., Lee, J., Cha, M. and Jang, H. (2017),” Predictability of machine learning techniques to forecast the trends of market index prices: Hypothesis testing for the Korean stock markets”, PloS one, Vol.12, No.11: e0188107.