A novel deep learning framework: Prediction and analysis of financial time series using CEEMD and LSTM

• Published in: Expert systems with applications Vol. 159; p. 113609
• Main Authors: Zhang, Yong'an, Yan, Binbin, Aasma, Memon
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 30.11.2020; Elsevier BV
• Subjects: Benchmarks; Complementary ensemble empirical mode decomposition; Decomposition; Deep learning; Empirical analysis; Feature extraction; Financial time series; Long short-term memory; Model accuracy; Model testing; Prediction models; Principal component analysis; Profitability; Robustness; Stock market forecasting; Time series; Deep learning; Long short-term memory; Stock market forecasting; Financial time series; Complementary ensemble empirical mode decomposition; Principal component analysis
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2020.113609

•A novel FTS forecasting methodology based on deep learning is proposed.•Proposed model exhibits highest predictive accuracy and directional symmetry.•Deep learning hybrid strategy yields stable excess returns and avoids drawdown risk.•Test error indicators generally drop as the stock markets maturity degree increases.•Methodology goes beyond a pure financial market application. Deep learning is well-known for extracting high-level abstract features from a large amount of raw data without relying on prior knowledge, which is potentially attractive in forecasting financial time series. Long short-term memory (LSTM) networks are deemed as state-of-the-art techniques in sequence learning, which are less commonly applied to financial time series predictions, yet inherently suitable for this domain. We propose a novel methodology of deep learning prediction, and based on this, construct a deep learning hybrid prediction model for stock markets—CEEMD-PCA-LSTM. In this model, complementary ensemble empirical mode decomposition (CEEMD), as a sequence smoothing and decomposition module, can decompose the fluctuations or trends of different scales of time series step by step, generating a series of intrinsic mode functions (IMFs) with different characteristic scales. Then, with retaining the most of information on raw data, PCA reduces dimension of the decomposed IMFs component, eliminating the redundant information and improving prediction response speed. After that, high-level abstract features are separately fed into LSTM networks to predict closing price of the next trading day for each component. Finally, synthesizing the predicted values of individual components is utilized to obtain a final predicted value. The empirical results of six representative stock indices from three types of markets indicate that our proposed model outperforms benchmark models in terms of predictive accuracy, i.e., lower test error and higher directional symmetry. Leveraging key research findings, we perform trading simulations to validate that the proposed model outperforms benchmark models in both absolute profitability performance and risk-adjusted profitability performance. Furthermore, model robustness test unveils the more stable robustness compared to benchmark models.