Author(s)

ZheShuo Zhang

Affiliation(s)

Wenzhou-Kean University, Wenzhou 325060, Zhejiang, China.

Corresponding Author

ZheShuo Zhang

ABSTRACT

This paper presents a comparative analysis of the effectiveness of hybrid CEEMDAN-LSTM models and traditional LSTM models in predicting realized volatility in financial markets. By utilizing realized volatility data from 2004 to 2024, the study highlights significant market fluctuations during the 2008 financial crisis and the 2020 COVID-19 pandemic. The findings indicate that the CEEMDAN-LSTM model, which decomposes time series data into intrinsic mode functions (IMFs) before applying LSTM networks, outperforms the basic LSTM model in terms of predictive accuracy, particularly during periods of high volatility. This enhanced performance is evidenced by lower error metrics, such as Mean Absolute Error (MAE) and Mean Squared Error (MSE). The research underscores the value of integrating advanced decomposition techniques with deep learning models to better capture the complex dynamics of financial markets.

KEYWORDS

CEEMDAN-LSTM; Intrinsic mode functions (IMFs); Mean Absolute Error (MAE); Mean Squared Error (MSE)

CITE THIS PAPER

ZheShuo Zhang. Performance comparison of CEEMDAN-LSTM and basic LSTM models in predicting realized volatility. Multidisciplinary Journal of Engineering and Technology. 2024, 1(1): 34-38. DOI: 10.61784/mjet3007.

REFERENCES

[1] Adebiyi A A, Adewumi A O, Ayo C K. Comparison of ARIMA and Artificial Neural Networks Models for Stock Price Prediction. Journal of Applied Mathematics, 2014: 1–7.

[2] Ak?ehir Z D, K?l?? E. A new denoising approach based on mode decomposition applied to the stock market time series: 2LE-CEEMDAN. PeerJ. Computer Science, 2024, 10, e1852.

[3] Cao J, Li Z, Li J. Financial time series forecasting model based on CEEMDAN and LSTM. Physica A: Statistical Mechanics and Its Applications, 2019, 519: 127–139.

[4] Guan Y. Financial time series forecasting model based on CEEMDAN-LSTM. 2022 4th International Conference on Advances in Computer Technology, Information Science and Communications (CTISC), 2022.

[5] Guresen E, Kayakutlu G, Daim T U. Using artificial neural network models in stock market index prediction. Expert Systems with Applications, 2011, 38(8): 10389–10397.

[6] Lin Y, Yan Y, Xu J, Liao Y, Ma F. Forecasting stock index price using the CEEMDAN-LSTM model. The North American Journal of Economics and Finance, 2021, 57: 101421.

[7] Lin Y, Lin Z, Liao Y, Li Y, Xu J, Yan Y. Forecasting the realized volatility of stock price index: A hybrid model integrating CEEMDAN and LSTM. Expert Systems with Applications, 2022, 206: 117736.

[8] Assaad M, Boné R, Cardot H. A new boosting algorithm for improved time-series forecasting with recurrent neural networks. Information Fusion, 2008, 9(1): 41–55.

[9] Baek Y, Kim H Y. ModAugNet: A new forecasting framework for stock market index value with an overfitting prevention LSTM module and a prediction LSTM module. Expert Systems with Applications, 2018, 113: 457–480.

[10] Pin Lv, Shu Y, Xu J, Wu Q. Modal decomposition-based hybrid model for stock index prediction. Expert Systems with Applications, 2022, 202: 117252–117252.

[11] Qi C, Ren J, Su J. GRU Neural Network Based on CEEMDAN–Wavelet for Stock Price Prediction. Applied Sciences, 2023, 13(12): 7104–7104.

[12] Sun Y, Liu J. AQI Prediction Based on CEEMDAN-ARMA-LSTM. Sustainability, 2022, 14(19): 12182.

[13] Wang J, Tang J, Guo K. Green Bond Index Prediction Based on CEEMDAN-LSTM. Frontiers in Energy Research, 2022, 9.

[14] Yanan G, Xiaoqun C, Bainian L, Kecheng P. Chaotic Time Series Prediction Using LSTM with CEEMDAN. Journal of Physics: Conference Series, 2020, 1617: 012094.