A decomposition-guided mechanism for nonstationary time series forecasting
Time series forecasting has been playing an important role in decision making, control, and monitoring across various fields. Specifically, the forecasting of nonstationarity time series remains a challenging problem where traditional time series modeling may not fully capture temporal dynamics. Rec...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIP Publishing LLC
2024-01-01
|
| Series: | AIP Advances |
| Online Access: | http://dx.doi.org/10.1063/5.0153647 |
| _version_ | 1797331965199777792 |
|---|---|
| author | Hao Wang Lubna Al Tarawneh Changqing Cheng Yu Jin |
| author_facet | Hao Wang Lubna Al Tarawneh Changqing Cheng Yu Jin |
| author_sort | Hao Wang |
| collection | DOAJ |
| description | Time series forecasting has been playing an important role in decision making, control, and monitoring across various fields. Specifically, the forecasting of nonstationarity time series remains a challenging problem where traditional time series modeling may not fully capture temporal dynamics. Recent studies of applying machine learning (ML) or more advanced hybrid models combining the ML and decomposition methods have shown their flexible nonstationary and nonlinear modeling capability. However, the end-effect problem introduced by the decomposition methods still introduces significant forecasting errors because of the unknown realm beyond the time series boundary. Therefore, a novel method applying a decomposition-guided mechanism is proposed in this work to eliminate the end effect problem while inheriting the knowledge learned from the decomposition state space to improve the prediction accuracy of such hybrid models in time series forecasting. Additionally, a domain adaptation model is integrated with the proposed mechanism to transfer knowledge from the source domain to the target domain regarding the decomposition state space. In this work, the intrinsic time-scale decomposition and Gaussian process are considered as examples of decomposition and ML methods to demonstrate the proposed mechanism’s effectiveness. Both simulation experiments and real-world case studies are conducted to show that a hybrid model with the proposed mechanism outperforms the conventional time series forecasting model, pure ML, and the original hybrid model in terms of prediction accuracy. |
| first_indexed | 2024-03-08T07:42:52Z |
| format | Article |
| id | doaj.art-634b59de55074a2c9f5b29b68f86feca |
| institution | Directory Open Access Journal |
| issn | 2158-3226 |
| language | English |
| last_indexed | 2024-03-08T07:42:52Z |
| publishDate | 2024-01-01 |
| publisher | AIP Publishing LLC |
| record_format | Article |
| series | AIP Advances |
| spelling | doaj.art-634b59de55074a2c9f5b29b68f86feca2024-02-02T16:46:07ZengAIP Publishing LLCAIP Advances2158-32262024-01-01141015254015254-1210.1063/5.0153647A decomposition-guided mechanism for nonstationary time series forecastingHao Wang0Lubna Al Tarawneh1Changqing Cheng2Yu Jin3Department of Systems Science and Industrial Engineering, State University of New York at Binghamton, Binghamton, New York 13902, USADepartment of Systems Science and Industrial Engineering, State University of New York at Binghamton, Binghamton, New York 13902, USADepartment of Systems Science and Industrial Engineering, State University of New York at Binghamton, Binghamton, New York 13902, USADepartment of Systems Science and Industrial Engineering, State University of New York at Binghamton, Binghamton, New York 13902, USATime series forecasting has been playing an important role in decision making, control, and monitoring across various fields. Specifically, the forecasting of nonstationarity time series remains a challenging problem where traditional time series modeling may not fully capture temporal dynamics. Recent studies of applying machine learning (ML) or more advanced hybrid models combining the ML and decomposition methods have shown their flexible nonstationary and nonlinear modeling capability. However, the end-effect problem introduced by the decomposition methods still introduces significant forecasting errors because of the unknown realm beyond the time series boundary. Therefore, a novel method applying a decomposition-guided mechanism is proposed in this work to eliminate the end effect problem while inheriting the knowledge learned from the decomposition state space to improve the prediction accuracy of such hybrid models in time series forecasting. Additionally, a domain adaptation model is integrated with the proposed mechanism to transfer knowledge from the source domain to the target domain regarding the decomposition state space. In this work, the intrinsic time-scale decomposition and Gaussian process are considered as examples of decomposition and ML methods to demonstrate the proposed mechanism’s effectiveness. Both simulation experiments and real-world case studies are conducted to show that a hybrid model with the proposed mechanism outperforms the conventional time series forecasting model, pure ML, and the original hybrid model in terms of prediction accuracy.http://dx.doi.org/10.1063/5.0153647 |
| spellingShingle | Hao Wang Lubna Al Tarawneh Changqing Cheng Yu Jin A decomposition-guided mechanism for nonstationary time series forecasting AIP Advances |
| title | A decomposition-guided mechanism for nonstationary time series forecasting |
| title_full | A decomposition-guided mechanism for nonstationary time series forecasting |
| title_fullStr | A decomposition-guided mechanism for nonstationary time series forecasting |
| title_full_unstemmed | A decomposition-guided mechanism for nonstationary time series forecasting |
| title_short | A decomposition-guided mechanism for nonstationary time series forecasting |
| title_sort | decomposition guided mechanism for nonstationary time series forecasting |
| url | http://dx.doi.org/10.1063/5.0153647 |
| work_keys_str_mv | AT haowang adecompositionguidedmechanismfornonstationarytimeseriesforecasting AT lubnaaltarawneh adecompositionguidedmechanismfornonstationarytimeseriesforecasting AT changqingcheng adecompositionguidedmechanismfornonstationarytimeseriesforecasting AT yujin adecompositionguidedmechanismfornonstationarytimeseriesforecasting AT haowang decompositionguidedmechanismfornonstationarytimeseriesforecasting AT lubnaaltarawneh decompositionguidedmechanismfornonstationarytimeseriesforecasting AT changqingcheng decompositionguidedmechanismfornonstationarytimeseriesforecasting AT yujin decompositionguidedmechanismfornonstationarytimeseriesforecasting |