Predicting Natural Gas Prices Based on a Novel Hybrid Model with Variational Mode Decomposition

Predicting carbon and oil price returns using hybrid models based on machine and deep learning

“…Furthermore, ARCH‐GARCH models and their extensions are assembled with empirical mode decomposition, neural networks and SVM (Abdollahi, 2020; Huang et al, 2021; Li & Lu, 2015; Zhang et al, 2018). On the other hand, artificial neural networks and decomposition algorithms have been recurrently employed in literature (Li et al, 2019; Liang et al, 2022; Lu et al, 2023; Sun et al, 2016; Sun & Huang, 2020; Zhu, 2012; Zhu et al, 2016; Zhu et al, 2019). In addition, neural networks have also been combined with adaptive neuro‐fuzzy systems and other types of neural networks (Atsalakis, 2016; Zhao et al, 2018).…”

“…On the other hand, machine learning techniques (Chiroma et al, 2015; Ramyar & Kianfar, 2019; Xu et al, 2020; Zhu et al, 2017) have been also employed. A promising methodology for accurately predicting carbon prices is the hybrid models that combine both conventional econometric models and machine learning techniques (Abdollahi & Ebrahimi, 2020; Atsalakis, 2016; Li & Lu, 2015; Liang et al, 2022; Lin et al, 2022; Lu et al, 2023; Wang et al, 2021; Zhang et al, 2018; Zhang et al, 2022; Zhu, 2012; Zhu et al, 2019). While traditional econometric models yield accurate results, they are inadequate in capturing the inherent non‐linear dynamics of carbon prices.…”

“…While traditional econometric models yield accurate results, they are inadequate in capturing the inherent non‐linear dynamics of carbon prices. Consequently, machine learning techniques have become an interesting alternative to cope with this limitation (Lu et al, 2023; Zhu et al, 2018). On the other hand, to overcome the drawbacks of single models and improve the carbon price prediction accuracy, hybrid models have been proposed to obtain better results (Liang et al, 2022; Lu et al, 2023).…”

“…Consequently, machine learning techniques have become an interesting alternative to cope with this limitation (Lu et al, 2023; Zhu et al, 2018). On the other hand, to overcome the drawbacks of single models and improve the carbon price prediction accuracy, hybrid models have been proposed to obtain better results (Liang et al, 2022; Lu et al, 2023). This is due to the fact that hybrid models incorporate the advantages of traditional econometric approaches and machine learning techniques, allowing for the capture of linear and non‐linear effects in time series (Tsay & Chen, 2018).…”