This code is the implementation of this paper (An Accurate GRU-Based Power Time-Series Prediction Approach With Selective State Updating and Stochastic Optimization).
TensorFlow-gpu = 1.14 Keras = 2.1.5
#Usage You can run the wind-dataset-hyper-parameters-select-GRU-SSU-AMG.py file directly to implement the wind power forecasting task.
Firstly, you need to place these two files (optimizers.py & recurrent.py) in the location specified by Keras (./anaconda3/envs/tf_1.14/Lib/site-packages/keras/).
Then, you can execute the following command:
python wind-dataset-hyper-parameters-select-GRU-SSU-AMG.py
If you are interested, please cite this paper.
@article{zhengaccurate, title={An Accurate GRU-Based Power Time-Series Prediction Approach With Selective State Updating and Stochastic Optimization}, author={Zheng, Wendong and Chen, Gang}, journal={IEEE Transactions on Cybernetics}, publisher={IEEE}, year= {2021}, doi= {10.1109/TCYB.2021.3121312}, }