Jul 7, 2019 · ES-RNN is a hybrid between classical state space forecasting models and modern RNNs that achieved a 9.4% sMAPE improvement in the M4 competition.

A GPU-enabled version of the hybrid ES-RNN model by Slawek et al that won the M4 time-series forecasting competition by a large margin. The details of our ...

Jul 29, 2019 · PreprintPDF Available. Fast ES-RNN: A GPU Implementation of the ES-RNN Algorithm. July 2019. July 2019. Authors: Andrew Redd · Andrew Redd.

Jul 7, 2019 · By vectorizing the original ES-RNN implementation and porting the algorithm to a GPU, this work achieves up to 322x training speedup ...

Jul 7, 2019 · A Distance Correlation-Based Approach to Characterize the Effectiveness of Recurrent Neural Networks for Time Series Forecasting. Time series ...

Fast ES-RNN: A GPU Implementation of the ES-RNN Algorithm. Go to Project Site · deep-learning forecasting · ← Predicting if a Kiva Loan Posting Will Expire.

Aug 16, 2019 · Variable length implementation would require us to use dynamic RNNs and masking for the different loss. It takes some work but we plan to add ...

Fast ES-RNN: A GPU Implementation of the ES-RNN Algorithm · 4 code implementations • 7 Jul 2019 • Andrew Redd, Kaung Khin, Aldo Marini.

Jun 21, 2018 · Fast ES-RNN: A GPU Implementation of the ES-RNN Algorithm. Preprint, 2019. PDF. Kaung Khin, Philipp Burckhardt, Rema Padman. A Deep Learning ...

Fast ES-RNN: A GPU Implementation of the ES-RNN Algorithm · 4 code implementations • 7 Jul 2019 • Andrew Redd, Kaung Khin, Aldo Marini.