The code in this repository implements a variety of ensemble learning methods applied to time series prediction.
The multi-objective optimization algorithms include MODE,NSGA2 and MOPSO. In addition to the above multi-objective optimization algorithms, some single-objective meta-heuristics are also used to achieve ensemble learning. Besides, we have updated the ensemble learning approach based on reinforcement learning, mainly including Q-learning and Sarsa.
These codes are related to 《A new hybrid ensemble deep reinforcement learning model for wind speed short term forecasting》 and 《A multi-factor driven spatiotemporal wind power prediction model based on ensemble deep graph attention reinforcement learning networks》.
For business reasons, we are unable to open source all of our code and data. The open source code has been modified to some extent to comply with the relevant regulations.
This code is mainly open source about the application of multi-objective optimization, single-objective optimization and reinforcement learning in the field of ensemble learning and time series prediction.
If the code is helpful to you, please cite the following paper：
@article{liu2020new,
title={A new hybrid ensemble deep reinforcement learning model for wind speed short term forecasting},
author={Liu, Hui and Yu, Chengqing and Wu, Haiping and Duan, Zhu and Yan, Guangxi},
journal={Energy},
volume={202},
pages={117794},
year={2020},
publisher={Elsevier}
}@article{chengqing2023multi,
title={A multi-factor driven spatiotemporal wind power prediction model based on ensemble deep graph attention reinforcement learning networks},
author={Chengqing, Yu and Guangxi, Yan and Chengming, Yu and Yu, Zhang and Xiwei, Mi},
journal={Energy},
volume={263},
pages={126034},
year={2023},
publisher={Elsevier}
}