Enhancing agricultural sustainability: Time series forecasting with ICEEMDAN-VMD-GRU for economic-resilience
In this study, we offer a dual-decomposition hybrid time series forecasting model that combines improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and variational mode decomposition (VMD) with gated recurrent unit (GRU) neural networks. By decomposing the non-stat...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-09-01
|
| Series: | Results in Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2590123025024934 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In this study, we offer a dual-decomposition hybrid time series forecasting model that combines improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and variational mode decomposition (VMD) with gated recurrent unit (GRU) neural networks. By decomposing the non-stationary input time series with ICEEMDAN first and then VMD, we extract intrinsic mode functions (IMFs) and oscillatory modes, which enables our model to capture both linear and nonlinear patterns more effectively, thereby addressing the challenges posed by complex, non-stationary patterns in time series data, leading to improved forecasting accuracy via GRU. Validation on two real-world datasets shows that the proposed model achieves significant improvements over baseline and other pre-existing combination methods such as ICEEMDAN-RNN, ICEEMDAN-LSTM, ICEEMDAN-GRU, VMD-RNN, VMD-LSTM, VMD-GRU, ICEEMDAN-VMD-RNN, and ICEEMDAN-VMD-LSTM. This is evidenced by key performance metrics namely, RMSE, MAPE, MAE, R2 and quantile loss. A reliable methodology for increasing the accuracy of time series forecasting in real agricultural datasets, particularly, the fruits of UK, is provided by the ICEEMDAN-VMD-GRU framework. |
|---|---|
| ISSN: | 2590-1230 |