2 code implementations • 9 May 2021 • Marcus Carpenter, Chunbo Luo, Xiao-Si Wang
Applying Dropout to a GRU model trained on only 28 days of data reduced the RMSE by 23%.
Time Series Time Series Forecasting