Sains Malaysiana 53(9)(2024): 3229-3239

http://doi.org/10.17576/jsm-2024-5309-25

 

Enhanced Foreign Exchange Volatility Forecasting using CEEMDAN with Optuna-Optimized Ensemble Deep Learning Model

(Ramalan Kemeruapan Tukaran Asing yang Dipertingkatkan menggunakan CEEMDAN dengan Model Pembelajaran Mendalam Ensembel Dioptimumkan Optuna)

 

REHAN KAUSAR1, FARHAT IQBAL2,3,*, ABDUL RAZIQ2, NAVEED SHEIKH4 & ABDUL REHMAN4

 

1Department of Statistics, Sardar Bahadur Khan Women's University, Quetta, Pakistan

2Department of Statistics, University of Balochistan, Quetta, Pakistan

3Department of Mathematics, Imam Abdulrahman Bin Faisal University, Saudi Arabia

4Department of Mathematics, University of Balochistan, Quetta, Pakistan

 

Diserahkan: 9 Disember 2023/Diterima: 15 Julai 2024

 

Abstract

Foreign Exchange (FX) is the largest financial market in the world, with a daily trading volume that significantly exceeds that of stock and futures markets. The prediction of FX volatility is a critical financial concern that has garnered significant attention from researchers and practitioners due to its far-reaching implications in the financial markets. This paper presents a novel hybrid ensemble forecasting model integrating a decomposition strategy and three deep learning (DL) models: Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and Convolutional Neural Network (CNN). This combination addresses individual models' limitations and further improves the accuracy and stability of FX volatility forecasting. The proposed approach utilizes the CEEMDAN technique to decompose volatility into multiple distinct intrinsic mode functions (IMFs) and merges these IMFs with GARCH and EGARCH volatilities to form the input dataset for the DL models. In addition, we employed an attention mechanism to improve the effectiveness of the DL techniques. Furthermore, the hyperparameters for the DL models are optimized using the Optuna algorithm. Finally, a hybrid ensemble model for forecasting exchange rate volatility is developed by combining the predictions of three distinct DL models. The proposed approach is evaluated against various benchmark models using evaluation measures such as MSE, MAE, HMSE, HMAE, RMSE, Q-LIKE, and the model confidence set (MCS) approach. The results demonstrate that our proposed approach provides accurate and reliable forecasts of FX volatility under different forecasting regimes, making it a valuable tool for financial practitioners and researchers.

 

Keywords: Currency exchange rate volatility; deep learning; ensemble; CEEMDAN; Optuna

 

Abstrak

Tukaran Asing (FX) merupakan pasaran kewangan terbesar di dunia dengan volum dagangan harian yang jauh melebihi pasaran saham dan pasaran hadapan. Ramalan turun naik FX merupakan kebimbangan kewangan yang kritikal serta telah mendapat perhatian daripada penyelidik dan pengamal kerana implikasinya yang meluas dalam pasaran kewangan. Kajian ini membentangkan model ramalan ensembel hibrid baharu yang menyepadukan strategi penguraian dan tiga model pembelajaran mendalam (DL): Memori Jangka Pendek Panjang (LSTM), LSTM Dwiarah (BiLSTM) dan Rangkaian Neural Konvolusi (CNN). Gabungan ini menangani had model individu dan meningkatkan lagi ketepatan dan kestabilan ramalan turun naik FX. Pendekatan yang dicadangkan menggunakan teknik CEEMDAN untuk menguraikan turun naik kepada pelbagai fungsi mod intrinsik (IMF) yang berbeza dan menggabungkan IMF ini dengan turun naik GARCH dan EGARCH untuk membentuk set data input bagi model DL. Di samping itu, kami menggunakan mekanisme perhatian untuk meningkatkan keberkesanan teknik DL. Tambahan pula, hiperparameter untuk model DL dioptimumkan menggunakan algoritma Optuna. Akhir sekali, model ensembel hibrid untuk meramalkan turun naik kadar pertukaran dibangunkan dengan menggabungkan ramalan tiga model DL yang berbeza. Pendekatan yang dicadangkan dinilai berdasarkan pelbagai model penanda aras menggunakan ukuran penilaian seperti MSE, MAE, HMSE, HMAE, RMSE, Q-like dan pendekatan set keyakinan model (MCS). Keputusan menunjukkan bahawa pendekatan yang dicadangkan dalam kajian ini menyediakan ramalan turun naik FX yang tepat dan boleh dipercayai di bawah rejim ramalan yang berbeza, menjadikannya alat yang berharga untuk pengamal dan penyelidik kewangan.

 

Kata kunci: CEEMDAN; ensembel; kemeruapan kadar pertukaran mata wang; Optuna; pembelajaran mendalam

 

RUJUKAN

Abedin, M.Z., Moon, M.H., Hassan, M.K. & Hajek, P. 2021. Deep learning-based exchange rate prediction during the COVID-19 pandemic. Annals of Operations Research. https://doi.org/10.1007/s10479-021-04420-6

Akiba, T., Sano, S., Yanase, T., Ohta, T. & Koyama, M. 2019. Optuna: A next-generation hyperparameter optimization framework. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. pp. 2623-2631. https://doi.org/10.1145/3292500.3330701

Baffour, A.A., Feng, J. & Taylor, E.K. 2019. A hybrid artificial neural network-GJR modeling approach to forecasting currency exchange rate volatility. Neurocomputing 365: 285-301. https://doi.org/10.1016/j.neucom.2019.07.088

Bollerslev, T. 1986. Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics 31(3): 307-327. https://doi.org/10.1016/0304-4076(86)90063-1

Borovkova, S. & Tsiamas, I. 2019. An ensemble of LSTM neural networks for high-frequency stock market classification. Journal of Forecasting 38(6): 600-619.  https://doi.org/10.1002/for.2585

Colominas, M.A., Schlotthauer, G. & Torres, M.E. 2014. Improved complete ensemble EMD: A suitable tool for biomedical signal processing. Biomedical Signal Processing and Control 14: 19-29. https://doi.org/10.1016/j.bspc.2014.06.009

Dauphin, Y.N., Fan, A., Auli, M. & Grangier, D. 2017. Language modeling with gated convolutional networks. ICML ’17: Proceedings of the 34th International Conference on Machine Learning 70: 933-941.

Dhamija, A.K. & Bhalla, V.K. 2010. Financial time series forecasting: Comparison of various arch models. Global Journal of Finance and Management 2(1): 159-172. http://www.ripublication.com/gjfm.htm

Diebold, F.X. & Mariano, R.S. 1995. Comparing predictive accuracy. Journal of Business and Economic Statistics 13(3): 253-263. https://doi.org/10.1198/073500102753410444

Dragomiretskiy, K. & Zosso, D. 2013. Variational mode decomposition. IEEE Transactions on Signal Processing 62(3): 531-544. https://doi.org/10.1109/TSP.2013.2288675

Engle, R.F. 1982. Autoregressive conditional heteroskedasticity with estimates of the variance of United Kingdom inflation. Econometrica 50(4): 987-1008.            https://doi.org/10.2307/1912773

Gamboa, D.A. 2017. A comparative study of volatility forecasting methods. The European Journal of Finance 23(7-9): 663-681.

Granata, F. & Di Nunno, F. 2021. Forecasting evapotranspiration in different climates using ensembles of recurrent neural networksAgricultural Water Management 255: 107040. https://doi.org/10.1016/j.agwat.2021.107040

Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., Liu, T., Wang, X., Wang, G., Cai, J. & Chen, T. 2018. Recent advances in convolutional neural networks. Pattern Recognition 77: 354-377. https://doi.org/10.1016/j.patcog.2017.10.013

Hajizadeh, E., Seifi, A., Zarandi, M.F. & Turksen, I.B. 2012. A hybrid modeling approach for forecasting the volatility of S&P 500 index return. Expert Systems with Applications 39(1): 431-436. https://doi.org/10.1016/j.eswa.2011.07.033

Hansen, P.R., Lunde, A. & Nason, J.M. 2011. The model confidence set. Econometrica 79(2): 453-497. https://doi.org/10.3982/ECTA5771

Henrique, B.M., Sobreiro, V.A. & Kimura, H. 2019. Literature review: Machine learning techniques applied to financial market prediction. Expert Systems with Applications 124: 226-251. https://doi.org/10.1016/j.eswa.2019.01.012 

Hochreiter, S. & Schmidhuber, J. 1997. Long short-term memory. Neural Computation 9(8): 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735

Iqbal, F., Zahid, M. & Koutmos, D. 2023. Cryptocurrency trading and downside risk. Risks 11(7): 122. https://doi.org/10.3390/risks11070122

Jung, G. & Choi, S.Y. 2021. Forecasting foreign exchange volatility using deep learning autoencoder-LSTM techniques. Complexity https://doi.org/10.1155/2021/6647534

Kalchbrenner, N., Espeholt, L., Simonyan, K., Oord, A.V.D., Graves, A. & Kavukcuoglu, K. 2016. Neural machine translation in linear time. ArXiv preprint ArXiv: 1610.10099.          https://doi.org/10.48550/arXiv.1610.10099

Kastner, G., Frühwirth-Schnatter, S. & Lopes, H.F. 2017. Efficient Bayesian inference for multivariate factor stochastic volatility models. Journal of Computational and Graphical Statistics 26(4): 905-917. https://doi.org/10.1080/10618600.2017.1322091

Kausar, R., Iqbal, F., Raziq, A. & Sheikh, N. 2023. A hybrid approach for accurate forecasting of exchange rate prices using VMD-CEEMDAN-GRU-ATCN model. Sains Malaysiana 52(11): 3293-3306. http://doi.org/10.17576/jsm-2023-5211-20

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. https://doi.org/10.1016/j.eswa.2018.03.002

Li, Y., Jiang, S., Li, X. & Wang, S. 2021. The role of news sentiment in oil futures returns and volatility forecasting: Data-decomposition based deep learning approach. Energy Economics 95: 105140. https://doi.org/10.1016/j.eneco.2021.105140

Pradeepkumar, D. & Ravi, V. 2017. Forecasting financial time series volatility using particle swarm optimization trained quantile regression neural network. Applied Soft Computing 58: 35-52. https://doi.org/10.1016/j.asoc.2017.04.014

Pradeepkumar, D. & Ravi, V. 2018. Soft computing hybrids for FOREX rate prediction: A comprehensive review. Computers & Operations Research 99: 262-284. https://doi.org/10.1016/j.cor.2018.05.020

Risse, M. 2019. Combining wavelet decomposition with machine learning to forecast gold returns. International Journal of Forecasting 35(2): 601-615. https://doi.org/10.1016/j.cor.2018.05.020

Santhosh, M., Venkaiah, C. & Kumar, D.V. 2019. Short-term wind speed forecasting approach using ensemble empirical mode decomposition and deep Boltzmann machine. Sustainable Energy, Grids and Networks 19: 100242. https://doi.org/10.1016/j.segan.2019.100242

Sobri, S., Koohi-Kamali, S. & Rahim, N.A. 2018. Solar photovoltaic generation forecasting methods: A review. Energy Conversion and Management 156: 459-497. https://doi.org/10.1016/j.enconman.2017.11.019

Taylor, S.J. 1994. Modeling stochastic volatility: A review and comparative study. Mathematical Finance 4(2): 183-204.  https://doi.org/10.1111/j.1467-9965.1994.tb00057.x

Torres, M.E., Colominas, M.A., Schlotthauer, G. & Flandrin, P. 2011. A complete ensemble empirical mode decomposition with adaptive noise. Proceedings of IEEE International Conference on Acoustics. https://10.1109/ICASSP.2011.5947265

Tripathi, M., Kumar, S. & Inani, S.K. 2020. Exchange rate forecasting using ensemble modeling for better policy implications. Journal of Time Series Econometrics 13(1): 43-71. https://doi.org/10.1515/jtse-2020-0013

Wu, Z.H. & Huang, N.E. 2009. Ensemble empirical mode decomposition: A noise-assisted data analysis method. Advances in Adaptive Data Analysis 01(01): 1-41. https://doi.org/10.1142/S1793536909000047

Zahid, M. & Iqbal, F. 2020. Modeling the volatility of cryptocurrencies: An empirical application of stochastic volatility models. Sains Malaysiana 49(3): 703-712. http://dx.doi.org/10.17576/jsm-2020-4903-25

Zahid, M., Iqbal, F. & Koutmos, D. 2022. Forecasting bitcoin volatility using hybrid GARCH models with machine learning. Risks 10(12): 237. https://doi.org/10.3390/risks10120237

 

*Pengarang untuk surat-menyurat; email: fsmuhammad@iau.edu.sa

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

sebelumnya